{ "updatedAt": "2026-06-04T03:03:53+00:00", "items": [ { "id": "paper:arxiv:2606.02800", "mode": "signal", "kind": "paper", "title": "Cosmos 3: Omnimodal World Models for Physical AI", "summary": "Hugging Face daily paper entry for Cosmos 3: Omnimodal World Models for Physical AI (2606.02800).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02800", "paper": "https://arxiv.org/abs/2606.02800", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.02060", "mode": "signal", "kind": "paper", "title": "Where Do Deep-Research Agents Go Wrong? Span-Level Error Localization in Agent Trajectories", "summary": "Hugging Face daily paper entry for Where Do Deep-Research Agents Go Wrong? Span-Level Error Localization in Agent Trajectories (2606.02060).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02060", "paper": "https://arxiv.org/abs/2606.02060", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03577", "mode": "signal", "kind": "paper", "title": "Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching", "summary": "Hugging Face daily paper entry for Eliciting Complex Spatial Reasoning in MLLMs through Wide-Baseline Matching (2606.03577).", "popularity": { "value": 4, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03577", "paper": "https://arxiv.org/abs/2606.03577", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01993", "mode": "signal", "kind": "paper", "title": "MMG2Skill: Can Agents Distill In-the-Wild Guides into Self-Evolving Skills?", "summary": "Hugging Face daily paper entry for MMG2Skill: Can Agents Distill In-the-Wild Guides into Self-Evolving Skills? (2606.01993).", "popularity": { "value": 4, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01993", "paper": "https://arxiv.org/abs/2606.01993", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04527", "mode": "signal", "kind": "paper", "title": "Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generation", "summary": "Hugging Face daily paper entry for Echo-Infinity: Learning Evolving Memory for Real-Time Infinite Video Generation (2606.04527).", "popularity": { "value": 3, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2606.04527", "paper": "https://arxiv.org/abs/2606.04527", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03031", "mode": "signal", "kind": "paper", "title": "AUDITFLOW: Executable Symbolic Environments for Structured Financial Reporting Verification", "summary": "Hugging Face daily paper entry for AUDITFLOW: Executable Symbolic Environments for Structured Financial Reporting Verification (2606.03031).", "popularity": { "value": 3, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03031", "paper": "https://arxiv.org/abs/2606.03031", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04513", "mode": "signal", "kind": "paper", "title": "MapAgent: An Industrial-Grade Agentic Framework for City-scale Lane-level Map Generation", "summary": "Hugging Face daily paper entry for MapAgent: An Industrial-Grade Agentic Framework for City-scale Lane-level Map Generation (2606.04513).", "popularity": { "value": 2, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.04513", "paper": "https://arxiv.org/abs/2606.04513", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03746", "mode": "signal", "kind": "paper", "title": "Qwen-Image-Flash: Beyond Objective Design", "summary": "Hugging Face daily paper entry for Qwen-Image-Flash: Beyond Objective Design (2606.03746).", "popularity": { "value": 2, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2606.03746", "paper": "https://arxiv.org/abs/2606.03746", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01286", "mode": "signal", "kind": "paper", "title": "BenchEvolver: Frontier Task Synthesis via Solution-Centric Evolution", "summary": "Hugging Face daily paper entry for BenchEvolver: Frontier Task Synthesis via Solution-Centric Evolution (2606.01286).", "popularity": { "value": 2, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01286", "paper": "https://arxiv.org/abs/2606.01286", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.05160", "mode": "signal", "kind": "paper", "title": "GRAIL: Generating Humanoid Loco-Manipulation from 3D Assets and Video Priors", "summary": "Hugging Face daily paper entry for GRAIL: Generating Humanoid Loco-Manipulation from 3D Assets and Video Priors (2606.05160).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2606.05160", "paper": "https://arxiv.org/abs/2606.05160", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.05158", "mode": "signal", "kind": "paper", "title": "Streaming Communication in Multi-Agent Reasoning", "summary": "Hugging Face daily paper entry for Streaming Communication in Multi-Agent Reasoning (2606.05158).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.05158", "paper": "https://arxiv.org/abs/2606.05158", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.05121", "mode": "signal", "kind": "paper", "title": "Audio Interaction Model", "summary": "Hugging Face daily paper entry for Audio Interaction Model (2606.05121).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.05121", "paper": "https://arxiv.org/abs/2606.05121", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.05008", "mode": "signal", "kind": "paper", "title": "M^3Eval: Multi-Modal Memory Evaluation through Cognitively-Grounded Video Tasks", "summary": "Hugging Face daily paper entry for M^3Eval: Multi-Modal Memory Evaluation through Cognitively-Grounded Video Tasks (2606.05008).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.05008", "paper": "https://arxiv.org/abs/2606.05008", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04036", "mode": "signal", "kind": "paper", "title": "Self-Distilled Policy Gradient", "summary": "Hugging Face daily paper entry for Self-Distilled Policy Gradient (2606.04036).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.04036", "paper": "https://arxiv.org/abs/2606.04036", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03197", "mode": "signal", "kind": "paper", "title": "MemTrain: Self-Supervised Context Memory Training", "summary": "Hugging Face daily paper entry for MemTrain: Self-Supervised Context Memory Training (2606.03197).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03197", "paper": "https://arxiv.org/abs/2606.03197", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01166", "mode": "signal", "kind": "paper", "title": "BraveGuard: From Open-World Threats to Safer Computer-Use Agents", "summary": "Hugging Face daily paper entry for BraveGuard: From Open-World Threats to Safer Computer-Use Agents (2606.01166).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01166", "paper": "https://arxiv.org/abs/2606.01166", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04688", "mode": "signal", "kind": "paper", "title": "MeshWeaver: Sparse-Voxel-Guided Surface Weaving for Autoregressive Mesh Generation", "summary": "Hugging Face daily paper entry for MeshWeaver: Sparse-Voxel-Guided Surface Weaving for Autoregressive Mesh Generation (2606.04688).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.04688", "paper": "https://arxiv.org/abs/2606.04688", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.00377", "mode": "signal", "kind": "paper", "title": "Score-Control for Hallucination Reduction in Diffusion Models", "summary": "Hugging Face daily paper entry for Score-Control for Hallucination Reduction in Diffusion Models (2606.00377).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-04", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00377", "paper": "https://arxiv.org/abs/2606.00377", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-04", "sourceUpdatedAt": "2026-06-04", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.00683", "mode": "signal", "kind": "paper", "title": "OCC-RAG: Optimal Cognitive Core for Faithful Question Answering", "summary": "Compact task-specialized language models demonstrate superior performance in multi-hop reasoning and faithfulness compared to larger general-purpose models through a novel training pipeline and structured reasoning traces. Recent progress in the development...", "popularity": { "value": 73, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00683", "paper": "https://arxiv.org/abs/2606.00683", "repo": "https://github.com/optimal-cognitive-core/OCC-RAG", "docs": "https://arxiv.org/pdf/2606.00683" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.23895", "mode": "signal", "kind": "paper", "title": "From Activation to Causality: Discovery of Causal Visual Representations in the Human Brain", "summary": "BrainCause framework uses generative and brain models to identify valid neural representations through causal testing, demonstrating that activation alone is insufficient for confirming concept representation. Identifying which brain regions represent a vis...", "popularity": { "value": 42, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23895", "paper": "https://arxiv.org/abs/2605.23895", "repo": "", "docs": "https://arxiv.org/pdf/2605.23895" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01249", "mode": "signal", "kind": "paper", "title": "Trust Region On-Policy Distillation", "summary": "Trust Region On-Policy Distillation (TrOPD) improves reliable token-level supervision in large language model distillation by using trust regions, outlier estimation, and off-policy guidance to address instability issues under distribution mismatch. On-Poli...", "popularity": { "value": 33, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01249", "paper": "https://arxiv.org/abs/2606.01249", "repo": "", "docs": "https://arxiv.org/pdf/2606.01249" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03985", "mode": "signal", "kind": "paper", "title": "Humanoid-GPT: Scaling Data and Structure for Zero-Shot Motion Tracking", "summary": "Humanoid-GPT is a GPT-style Transformer with causal attention trained on a billion-scale motion corpus that achieves zero-shot generalization to unseen motions and control tasks through scalable pre-training on diverse motion data. We introduce Humanoid-GPT...", "popularity": { "value": 32, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03985", "paper": "https://arxiv.org/abs/2606.03985", "repo": "", "docs": "https://arxiv.org/pdf/2606.03985" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03458", "mode": "signal", "kind": "paper", "title": "KVarN: Variance-Normalized KV-Cache Quantization Mitigates Error Accumulation in Reasoning Tasks", "summary": "KVarN is a calibration-free KV-cache quantizer that uses Hadamard rotation and dual-scaling variance normalization to reduce error accumulation during autoregressive decoding in large language models. Test-time scaling is a powerful approach to obtain bette...", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03458", "paper": "https://arxiv.org/abs/2606.03458", "repo": "https://github.com/huawei-csl/KVarN", "docs": "https://arxiv.org/pdf/2606.03458" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.02398", "mode": "signal", "kind": "paper", "title": "A Local Perturbation Theory for Cross-Domain Interference and Recovery in Multi-Domain RL", "summary": "Multi-domain reinforcement learning in language models causes performance degradation through shared computational pathways, but targeted refresh and rollback techniques can selectively recover lost capabilities with minimal side effects. Reinforcement lear...", "popularity": { "value": 24, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02398", "paper": "https://arxiv.org/abs/2606.02398", "repo": "", "docs": "https://arxiv.org/pdf/2606.02398" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03603", "mode": "signal", "kind": "paper", "title": "World Models Meet Language Models: On the Complementarity of Concrete and Abstract Reasoning", "summary": "Controlled concrete reasoning combines visual simulation with abstract reasoning through a training method that uses privileged future information to improve prediction accuracy and robustness. World models and multimodal large language models (MLLMs) provi...", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03603", "paper": "https://arxiv.org/abs/2606.03603", "repo": "https://github.com/yczhou001/PF-OPSD", "docs": "https://arxiv.org/pdf/2606.03603" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01961", "mode": "signal", "kind": "paper", "title": "AutoMedBench: Towards Medical AutoResearch with Agentic AI Models", "summary": "AutoMedBench presents a comprehensive benchmark for autonomous medical-AI research that evaluates agent performance across five workflow stages, revealing validation as the weakest stage and highlighting the importance of reliable pipeline execution and ver...", "popularity": { "value": 21, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01961", "paper": "https://arxiv.org/abs/2606.01961", "repo": "https://github.com/AutoMedBench/AutoMedBench", "docs": "https://arxiv.org/pdf/2606.01961" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.30288", "mode": "signal", "kind": "paper", "title": "MIRA: Mid-training Rubric Anchoring for Source-Aware Data Selection", "summary": "MIRA is a source-aware filtering framework for mid-training data selection in LLM development that uses self-anchored rubric discovery to balance scalability and semantic accuracy across heterogeneous data sources. Mid-training has become an important stage...", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30288", "paper": "https://arxiv.org/abs/2605.30288", "repo": "", "docs": "https://arxiv.org/pdf/2605.30288" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03979", "mode": "signal", "kind": "paper", "title": "Language Models Need Sleep: Learning to Self-Modify and Consolidate Memories", "summary": "Deep learning models with sleep and dreaming paradigms enable continual learning through memory consolidation and self-improvement phases. The past few decades have witnessed significant advances in the design of machine learning algorithms, from early stud...", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03979", "paper": "https://arxiv.org/abs/2606.03979", "repo": "https://github.com/lexasub/somnolent_llm/", "docs": "https://arxiv.org/pdf/2606.03979" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03920", "mode": "signal", "kind": "paper", "title": "Benchmarking Visual State Tracking in Multimodal Video Understanding", "summary": "Current multimodal large language models struggle with visual state tracking in videos, performing poorly even when human-level capabilities are required, and existing agentic approaches do not effectively address these limitations. Understanding a video re...", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03920", "paper": "https://arxiv.org/abs/2606.03920", "repo": "https://github.com/vision-x-nyu/vstat", "docs": "https://arxiv.org/pdf/2606.03920" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01599", "mode": "signal", "kind": "paper", "title": "TRON: Targeted Rule-Verifiable Online Environments for Visual Reasoning RL", "summary": "TRON enables scalable and controllable reinforcement learning for visual reasoning through an online environment substrate that generates unlimited diverse training instances with verifiable answers. Reinforcement learning (RL) for visual reasoning needs sc...", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01599", "paper": "https://arxiv.org/abs/2606.01599", "repo": "https://github.com/YangTianze009/TRON", "docs": "https://arxiv.org/pdf/2606.01599" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03159", "mode": "signal", "kind": "paper", "title": "NVIDIA OmniDreams: Real-Time Generative World Model for Closed-Loop Autonomous Vehicle Simulation", "summary": "OmniDreams, a foundation generative world model trained from the Cosmos diffusion model, enables real-time action-conditioned video generation for autonomous driving policy evaluation in complex, unseen scenarios. As autonomous vehicle capabilities advance,...", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03159", "paper": "https://arxiv.org/abs/2606.03159", "repo": "", "docs": "https://arxiv.org/pdf/2606.03159" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01717", "mode": "signal", "kind": "paper", "title": "Decentralized Instruction Tuning: Conflict-Aware Splitting and Weight Merging", "summary": "Instruction tuning of large language models can be improved through decentralized training that partitions mixed datasets based on gradient conflicts and merges results via weighted averaging, achieving performance comparable to centralized methods with red...", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01717", "paper": "https://arxiv.org/abs/2606.01717", "repo": "https://github.com/naver-ai/merit", "docs": "https://arxiv.org/pdf/2606.01717" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03911", "mode": "signal", "kind": "paper", "title": "Bootstrap Your Generator: Unpaired Visual Editing with Flow Matching", "summary": "Bootstrap Your Generator framework enables unpaired training of flow matching editing models by leveraging base model knowledge and gradient routing for improved generalization in data-scarce scenarios. Modern generative models possess a deep understanding...", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03911", "paper": "https://arxiv.org/abs/2606.03911", "repo": "", "docs": "https://arxiv.org/pdf/2606.03911" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.30039", "mode": "signal", "kind": "paper", "title": "Domain-Specific Data Synthesis for LLMs via Minimal Sufficient Representation Learning", "summary": "Hugging Face daily paper entry for Domain-Specific Data Synthesis for LLMs via Minimal Sufficient Representation Learning (2605.30039).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30039", "paper": "https://arxiv.org/abs/2605.30039", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.02754", "mode": "signal", "kind": "paper", "title": "Ψ-Bench: Evaluating Persona-Sensitive Influencing in Persuasive Dialogues", "summary": "Hugging Face daily paper entry for Ψ-Bench: Evaluating Persona-Sensitive Influencing in Persuasive Dialogues (2606.02754).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02754", "paper": "https://arxiv.org/abs/2606.02754", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01048", "mode": "signal", "kind": "paper", "title": "Decoupled Residual Denoising Diffusion Models for Unified and Data Efficient Image-to-Image Translation", "summary": "Hugging Face daily paper entry for Decoupled Residual Denoising Diffusion Models for Unified and Data Efficient Image-to-Image Translation (2606.01048).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2606.01048", "paper": "https://arxiv.org/abs/2606.01048", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03102", "mode": "signal", "kind": "paper", "title": "Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time Scaling", "summary": "Hugging Face daily paper entry for Small RL Controller, Large Language Model: RL-Guided Adaptive Sampling for Test-Time Scaling (2606.03102).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03102", "paper": "https://arxiv.org/abs/2606.03102", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01770", "mode": "signal", "kind": "paper", "title": "Adaptive Auto-Harness: Sustained Self-Improvement for Agentic System Deployment on Open-Ended Task Streams", "summary": "Hugging Face daily paper entry for Adaptive Auto-Harness: Sustained Self-Improvement for Agentic System Deployment on Open-Ended Task Streams (2606.01770).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01770", "paper": "https://arxiv.org/abs/2606.01770", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03264", "mode": "signal", "kind": "paper", "title": "PaddleOCR-VL-1.6: Expanding the Frontier of Document Parsing with Under-Optimized Region Refinement and Progressive Post-Training", "summary": "Hugging Face daily paper entry for PaddleOCR-VL-1.6: Expanding the Frontier of Document Parsing with Under-Optimized Region Refinement and Progressive Post-Training (2606.03264).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03264", "paper": "https://arxiv.org/abs/2606.03264", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01788", "mode": "signal", "kind": "paper", "title": "PlatonicNav: Unveiling Semantic Correspondence in Navigation with Platonic Topological Maps", "summary": "Hugging Face daily paper entry for PlatonicNav: Unveiling Semantic Correspondence in Navigation with Platonic Topological Maps (2606.01788).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01788", "paper": "https://arxiv.org/abs/2606.01788", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.29288", "mode": "signal", "kind": "paper", "title": "Diagnosing Harmful Continuation in Answer-Correct Long-CoT Training Traces", "summary": "Hugging Face daily paper entry for Diagnosing Harmful Continuation in Answer-Correct Long-CoT Training Traces (2605.29288).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29288", "paper": "https://arxiv.org/abs/2605.29288", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03928", "mode": "signal", "kind": "paper", "title": "Value-Aware Stochastic KV Cache Eviction for Reasoning Models", "summary": "Hugging Face daily paper entry for Value-Aware Stochastic KV Cache Eviction for Reasoning Models (2606.03928).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03928", "paper": "https://arxiv.org/abs/2606.03928", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01476", "mode": "signal", "kind": "paper", "title": "OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification", "summary": "Hugging Face daily paper entry for OmniOPD: Logit-Free On-Policy Distillation via Speculative Verification (2606.01476).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01476", "paper": "https://arxiv.org/abs/2606.01476", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.27346", "mode": "signal", "kind": "paper", "title": "MERIT: Learning Disentangled Music Representations for Audio Similarity", "summary": "Hugging Face daily paper entry for MERIT: Learning Disentangled Music Representations for Audio Similarity (2605.27346).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27346", "paper": "https://arxiv.org/abs/2605.27346", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03748", "mode": "signal", "kind": "paper", "title": "Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models", "summary": "Hugging Face daily paper entry for Ultralytics YOLO26: Unified Real-Time End-to-End Vision Models (2606.03748).", "popularity": { "value": 4, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2606.03748", "paper": "https://arxiv.org/abs/2606.03748", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03029", "mode": "signal", "kind": "paper", "title": "Conditional Hypothesis Generation for LLM-Based Text Analysis with Researcher-Specified Covariates", "summary": "Hugging Face daily paper entry for Conditional Hypothesis Generation for LLM-Based Text Analysis with Researcher-Specified Covariates (2606.03029).", "popularity": { "value": 4, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03029", "paper": "https://arxiv.org/abs/2606.03029", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.00138", "mode": "signal", "kind": "paper", "title": "A Multi-AI-agent Framework Enabling End-to-end Finite Element Analysis for Solid Mechanics Problems", "summary": "Hugging Face daily paper entry for A Multi-AI-agent Framework Enabling End-to-end Finite Element Analysis for Solid Mechanics Problems (2606.00138).", "popularity": { "value": 4, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00138", "paper": "https://arxiv.org/abs/2606.00138", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01494", "mode": "signal", "kind": "paper", "title": "ClawHub Security Signals: When VirusTotal, Static Analysis, and SkillSpector Disagree", "summary": "Hugging Face daily paper entry for ClawHub Security Signals: When VirusTotal, Static Analysis, and SkillSpector Disagree (2606.01494).", "popularity": { "value": 3, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01494", "paper": "https://arxiv.org/abs/2606.01494", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.00386", "mode": "signal", "kind": "paper", "title": "αDepth: Learning Single-Pass Soft Boundary Decomposition for Stereo Conversion", "summary": "Hugging Face daily paper entry for αDepth: Learning Single-Pass Soft Boundary Decomposition for Stereo Conversion (2606.00386).", "popularity": { "value": 3, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00386", "paper": "https://arxiv.org/abs/2606.00386", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04433", "mode": "signal", "kind": "paper", "title": "Stateful Visual Encoders for Vision-Language Models", "summary": "Hugging Face daily paper entry for Stateful Visual Encoders for Vision-Language Models (2606.04433).", "popularity": { "value": 2, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.04433", "paper": "https://arxiv.org/abs/2606.04433", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.02775", "mode": "signal", "kind": "paper", "title": "AURA: Action-Gated Memory for Robot Policies at Constant VRAM", "summary": "Hugging Face daily paper entry for AURA: Action-Gated Memory for Robot Policies at Constant VRAM (2606.02775).", "popularity": { "value": 2, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02775", "paper": "https://arxiv.org/abs/2606.02775", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.30581", "mode": "signal", "kind": "paper", "title": "Prior Availability in Industrial Visual Sim-to-Real: A Review of CAD-Guided and CAD-Unavailable Regimes", "summary": "Hugging Face daily paper entry for Prior Availability in Industrial Visual Sim-to-Real: A Review of CAD-Guided and CAD-Unavailable Regimes (2605.30581).", "popularity": { "value": 2, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30581", "paper": "https://arxiv.org/abs/2605.30581", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04455", "mode": "signal", "kind": "paper", "title": "The Meta-Agent Challenge: Are Current Agents Capable of Autonomous Agent Development?", "summary": "Hugging Face daily paper entry for The Meta-Agent Challenge: Are Current Agents Capable of Autonomous Agent Development? (2606.04455).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.04455", "paper": "https://arxiv.org/abs/2606.04455", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03348", "mode": "signal", "kind": "paper", "title": "SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation", "summary": "Hugging Face daily paper entry for SynCred-Bench: Benchmarking Synthetic Credibility in AI-Generated Visual Misinformation (2606.03348).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03348", "paper": "https://arxiv.org/abs/2606.03348", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.03287", "mode": "signal", "kind": "paper", "title": "BA-T: An Iterative Transformer for Two-View Bundle Adjustment", "summary": "Hugging Face daily paper entry for BA-T: An Iterative Transformer for Two-View Bundle Adjustment (2606.03287).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.03287", "paper": "https://arxiv.org/abs/2606.03287", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.02578", "mode": "signal", "kind": "paper", "title": "Mitigating Perceptual Judgment Bias in Multimodal LLM-as-a-Judge via Perceptual Perturbation and Reward Modeling", "summary": "Hugging Face daily paper entry for Mitigating Perceptual Judgment Bias in Multimodal LLM-as-a-Judge via Perceptual Perturbation and Reward Modeling (2606.02578).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02578", "paper": "https://arxiv.org/abs/2606.02578", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.02461", "mode": "signal", "kind": "paper", "title": "AgentCL: Toward Rigorous Evaluation of Continual Learning in Language Agents", "summary": "Hugging Face daily paper entry for AgentCL: Toward Rigorous Evaluation of Continual Learning in Language Agents (2606.02461).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02461", "paper": "https://arxiv.org/abs/2606.02461", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.27958", "mode": "signal", "kind": "paper", "title": "Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations", "summary": "Hugging Face daily paper entry for Pressure-Testing Deception Probes in LLMs: Scaling, Robustness, and the Geometry of Deceptive Representations (2605.27958).", "popularity": { "value": 1, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27958", "paper": "https://arxiv.org/abs/2605.27958", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.04889", "mode": "signal", "kind": "paper", "title": "GRAIL: Gradient-Reweighted Advantages for Reinforcement Learning with Verifiable Rewards", "summary": "Hugging Face daily paper entry for GRAIL: Gradient-Reweighted Advantages for Reinforcement Learning with Verifiable Rewards (2606.04889).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.04889", "paper": "https://arxiv.org/abs/2606.04889", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2606.01955", "mode": "signal", "kind": "paper", "title": "WALL-WM: Carving World Action Modeling at the Event Joints", "summary": "Hugging Face daily paper entry for WALL-WM: Carving World Action Modeling at the Event Joints (2606.01955).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-03", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01955", "paper": "https://arxiv.org/abs/2606.01955", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-03", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.30611", "mode": "signal", "kind": "paper", "title": "Crafter: A Multi-Agent Harness for Editable Scientific Figure Generation from Diverse Inputs", "summary": "Hugging Face daily paper entry for Crafter: A Multi-Agent Harness for Editable Scientific Figure Generation from Diverse Inputs (2605.30611).", "popularity": { "value": 112, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30611", "paper": "https://arxiv.org/abs/2605.30611", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02437", "mode": "signal", "kind": "paper", "title": "On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters", "summary": "Hugging Face daily paper entry for On the Scaling of PEFT: Towards Million Personal Models of Trillion Parameters (2606.02437).", "popularity": { "value": 86, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02437", "paper": "https://arxiv.org/abs/2606.02437", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28556", "mode": "signal", "kind": "paper", "title": "A Matter of TASTE: Improving Coverage and Difficulty of Agent Benchmarks", "summary": "Hugging Face daily paper entry for A Matter of TASTE: Improving Coverage and Difficulty of Agent Benchmarks (2605.28556).", "popularity": { "value": 55, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28556", "paper": "https://arxiv.org/abs/2605.28556", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29707", "mode": "signal", "kind": "paper", "title": "Domino: Decoupling Causal Modeling from Autoregressive Drafting in Speculative Decoding", "summary": "Hugging Face daily paper entry for Domino: Decoupling Causal Modeling from Autoregressive Drafting in Speculative Decoding (2605.29707).", "popularity": { "value": 52, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29707", "paper": "https://arxiv.org/abs/2605.29707", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02404", "mode": "signal", "kind": "paper", "title": "K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean Contexts", "summary": "Hugging Face daily paper entry for K-BrowseComp: A Web Browsing Agent Benchmark Grounded in Korean Contexts (2606.02404).", "popularity": { "value": 47, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02404", "paper": "https://arxiv.org/abs/2606.02404", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.00408", "mode": "signal", "kind": "paper", "title": "Masking Stale Observations Helps Search Agents -- Until It Doesn't: A Regime Map and Its Mechanism", "summary": "Hugging Face daily paper entry for Masking Stale Observations Helps Search Agents -- Until It Doesn't: A Regime Map and Its Mechanism (2606.00408).", "popularity": { "value": 43, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00408", "paper": "https://arxiv.org/abs/2606.00408", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02373", "mode": "signal", "kind": "paper", "title": "Harness-1: Reinforcement Learning for Search Agents with State-Externalizing Harnesses", "summary": "Hugging Face daily paper entry for Harness-1: Reinforcement Learning for Search Agents with State-Externalizing Harnesses (2606.02373).", "popularity": { "value": 34, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02373", "paper": "https://arxiv.org/abs/2606.02373", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29343", "mode": "signal", "kind": "paper", "title": "Draft-OPD: On-Policy Distillation for Speculative Draft Models", "summary": "Hugging Face daily paper entry for Draft-OPD: On-Policy Distillation for Speculative Draft Models (2605.29343).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29343", "paper": "https://arxiv.org/abs/2605.29343", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24956", "mode": "signal", "kind": "paper", "title": "NITP: Next Implicit Token Prediction for LLM Pre-training", "summary": "Hugging Face daily paper entry for NITP: Next Implicit Token Prediction for LLM Pre-training (2605.24956).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24956", "paper": "https://arxiv.org/abs/2605.24956", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30501", "mode": "signal", "kind": "paper", "title": "Linear Ensembles Wash Away Watermarks: On the Fragility of Distributional Perturbations in LLMs", "summary": "Hugging Face daily paper entry for Linear Ensembles Wash Away Watermarks: On the Fragility of Distributional Perturbations in LLMs (2605.30501).", "popularity": { "value": 25, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30501", "paper": "https://arxiv.org/abs/2605.30501", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02482", "mode": "signal", "kind": "paper", "title": "X-Stream: Exploring MLLMs as Multiplexers for Multi-Stream Understanding", "summary": "Hugging Face daily paper entry for X-Stream: Exploring MLLMs as Multiplexers for Multi-Stream Understanding (2606.02482).", "popularity": { "value": 24, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02482", "paper": "https://arxiv.org/abs/2606.02482", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02564", "mode": "signal", "kind": "paper", "title": "VLMs are Good Teachers for Video Reasoning via Adaptive Test-Time Optimization", "summary": "Hugging Face daily paper entry for VLMs are Good Teachers for Video Reasoning via Adaptive Test-Time Optimization (2606.02564).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "multimodal", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2606.02564", "paper": "https://arxiv.org/abs/2606.02564", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30351", "mode": "signal", "kind": "paper", "title": "VideoMLA: Low-Rank Latent KV Cache for Minute-Scale Autoregressive Video Diffusion", "summary": "Hugging Face daily paper entry for VideoMLA: Low-Rank Latent KV Cache for Minute-Scale Autoregressive Video Diffusion (2605.30351).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30351", "paper": "https://arxiv.org/abs/2605.30351", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.01247", "mode": "signal", "kind": "paper", "title": "Where to Look: Can Foundation Models Reach a Target Viewpoint Through Active Exploration?", "summary": "Hugging Face daily paper entry for Where to Look: Can Foundation Models Reach a Target Viewpoint Through Active Exploration? (2606.01247).", "popularity": { "value": 21, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01247", "paper": "https://arxiv.org/abs/2606.01247", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.01311", "mode": "signal", "kind": "paper", "title": "SkillAdaptor: Self-Adapting Skills for LLM Agents from Trajectories", "summary": "Hugging Face daily paper entry for SkillAdaptor: Self-Adapting Skills for LLM Agents from Trajectories (2606.01311).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01311", "paper": "https://arxiv.org/abs/2606.01311", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28132", "mode": "signal", "kind": "paper", "title": "Which Pretraining Paradigm Better Serves Spatial Intelligence? An Empirical Comparison of Vision-Language and Video Generation Models", "summary": "Hugging Face daily paper entry for Which Pretraining Paradigm Better Serves Spatial Intelligence? An Empirical Comparison of Vision-Language and Video Generation Models (2605.28132).", "popularity": { "value": 18, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "multimodal", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.28132", "paper": "https://arxiv.org/abs/2605.28132", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29860", "mode": "signal", "kind": "paper", "title": "ESPO: Early-Stopping Proximal Policy Optimization", "summary": "Hugging Face daily paper entry for ESPO: Early-Stopping Proximal Policy Optimization (2605.29860).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29860", "paper": "https://arxiv.org/abs/2605.29860", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02470", "mode": "signal", "kind": "paper", "title": "MCP-Persona: Benchmarking LLM Agents on Real-World Personal Applications via Environment Simulation", "summary": "Hugging Face daily paper entry for MCP-Persona: Benchmarking LLM Agents on Real-World Personal Applications via Environment Simulation (2606.02470).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02470", "paper": "https://arxiv.org/abs/2606.02470", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24202", "mode": "signal", "kind": "paper", "title": "When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs", "summary": "Hugging Face daily paper entry for When Does Multi-Agent RL Improve LLM Workflows? Workflow, Scale, and Policy-Sharing Tradeoffs (2605.24202).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24202", "paper": "https://arxiv.org/abs/2605.24202", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31057", "mode": "signal", "kind": "paper", "title": "LVSA: Training-Free Sparse Attention for Long Video Diffusion", "summary": "Hugging Face daily paper entry for LVSA: Training-Free Sparse Attention for Long Video Diffusion (2605.31057).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.31057", "paper": "https://arxiv.org/abs/2605.31057", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02553", "mode": "signal", "kind": "paper", "title": "LongLive-RAG: A General Retrieval-Augmented Framework for Long Video Generation", "summary": "Hugging Face daily paper entry for LongLive-RAG: A General Retrieval-Augmented Framework for Long Video Generation (2606.02553).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02553", "paper": "https://arxiv.org/abs/2606.02553", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.01528", "mode": "signal", "kind": "paper", "title": "Joint Agent Memory and Exploration Learning via Novelty Signals", "summary": "Hugging Face daily paper entry for Joint Agent Memory and Exploration Learning via Novelty Signals (2606.01528).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01528", "paper": "https://arxiv.org/abs/2606.01528", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29588", "mode": "signal", "kind": "paper", "title": "Brain-IT-VQA: From Brain Signals to Answers", "summary": "Hugging Face daily paper entry for Brain-IT-VQA: From Brain Signals to Answers (2605.29588).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29588", "paper": "https://arxiv.org/abs/2605.29588", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25659", "mode": "signal", "kind": "paper", "title": "StreamChar: Long-Horizon Streaming Character Audio-Video Generation with Decoupled Orchestration", "summary": "Hugging Face daily paper entry for StreamChar: Long-Horizon Streaming Character Audio-Video Generation with Decoupled Orchestration (2605.25659).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.25659", "paper": "https://arxiv.org/abs/2605.25659", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02031", "mode": "signal", "kind": "paper", "title": "OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web Agents", "summary": "Hugging Face daily paper entry for OpenWebRL: Demystifying Online Multi-turn Reinforcement Learning for Visual Web Agents (2606.02031).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02031", "paper": "https://arxiv.org/abs/2606.02031", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30723", "mode": "signal", "kind": "paper", "title": "Skill is Not One-Size-Fits-All: Model-Aware Skill Alignment for LLM Agents", "summary": "Hugging Face daily paper entry for Skill is Not One-Size-Fits-All: Model-Aware Skill Alignment for LLM Agents (2605.30723).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30723", "paper": "https://arxiv.org/abs/2605.30723", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02388", "mode": "signal", "kind": "paper", "title": "Policy and World Modeling Co-Training for Language Agents", "summary": "Hugging Face daily paper entry for Policy and World Modeling Co-Training for Language Agents (2606.02388).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02388", "paper": "https://arxiv.org/abs/2606.02388", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30852", "mode": "signal", "kind": "paper", "title": "Speculative Pipeline Decoding: Higher-Accruacy and Zero-Bubble Speculation via Pipeline Parallelism", "summary": "Hugging Face daily paper entry for Speculative Pipeline Decoding: Higher-Accruacy and Zero-Bubble Speculation via Pipeline Parallelism (2605.30852).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30852", "paper": "https://arxiv.org/abs/2605.30852", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30931", "mode": "signal", "kind": "paper", "title": "MineExplorer: Evaluating Open-World Exploration of MLLM Agents in Minecraft", "summary": "Hugging Face daily paper entry for MineExplorer: Evaluating Open-World Exploration of MLLM Agents in Minecraft (2605.30931).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30931", "paper": "https://arxiv.org/abs/2605.30931", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02551", "mode": "signal", "kind": "paper", "title": "AFUN: Towards an Affordance Foundation Model for Functionality Understanding", "summary": "Hugging Face daily paper entry for AFUN: Towards an Affordance Foundation Model for Functionality Understanding (2606.02551).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02551", "paper": "https://arxiv.org/abs/2606.02551", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.01414", "mode": "signal", "kind": "paper", "title": "Agent Skills Should Go Beyond Text: The Case for Visual Skills", "summary": "Hugging Face daily paper entry for Agent Skills Should Go Beyond Text: The Case for Visual Skills (2606.01414).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01414", "paper": "https://arxiv.org/abs/2606.01414", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.00828", "mode": "signal", "kind": "paper", "title": "RoboStressBench: Benchmarking VLM Robustness to Physical Visual Stress in Embodied Scenes", "summary": "Hugging Face daily paper entry for RoboStressBench: Benchmarking VLM Robustness to Physical Visual Stress in Embodied Scenes (2606.00828).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00828", "paper": "https://arxiv.org/abs/2606.00828", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30126", "mode": "signal", "kind": "paper", "title": "PARCEL: Pool-Anchored Resampling with Conditioned Elastic Queries for Efficient Vision-Language Understanding", "summary": "Hugging Face daily paper entry for PARCEL: Pool-Anchored Resampling with Conditioned Elastic Queries for Efficient Vision-Language Understanding (2605.30126).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.30126", "paper": "https://arxiv.org/abs/2605.30126", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02277", "mode": "signal", "kind": "paper", "title": "RoboSemanticBench: Diagnosing Semantic Grounding in Action Prediction for VLA Models", "summary": "Hugging Face daily paper entry for RoboSemanticBench: Diagnosing Semantic Grounding in Action Prediction for VLA Models (2606.02277).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02277", "paper": "https://arxiv.org/abs/2606.02277", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.01682", "mode": "signal", "kind": "paper", "title": "Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning", "summary": "Hugging Face daily paper entry for Off-the-Shelf LLMs as Process Scorers: Training-Free Alternative to PRMs for Mathematical Reasoning (2606.01682).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01682", "paper": "https://arxiv.org/abs/2606.01682", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.01533", "mode": "signal", "kind": "paper", "title": "Multi-Agent Computer Use", "summary": "Hugging Face daily paper entry for Multi-Agent Computer Use (2606.01533).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01533", "paper": "https://arxiv.org/abs/2606.01533", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31597", "mode": "signal", "kind": "paper", "title": "SOCO: Benchmarking Semantic Object Correspondence in Vision Foundation Models", "summary": "Hugging Face daily paper entry for SOCO: Benchmarking Semantic Object Correspondence in Vision Foundation Models (2605.31597).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31597", "paper": "https://arxiv.org/abs/2605.31597", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24614", "mode": "signal", "kind": "paper", "title": "Measuring the Depth of LLM Unlearning via Activation Patching", "summary": "Hugging Face daily paper entry for Measuring the Depth of LLM Unlearning via Activation Patching (2605.24614).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24614", "paper": "https://arxiv.org/abs/2605.24614", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2606.02580", "mode": "signal", "kind": "paper", "title": "Thinking in Blender: Staged Executable Inverse Graphics with Vision-Language Models", "summary": "Hugging Face daily paper entry for Thinking in Blender: Staged Executable Inverse Graphics with Vision-Language Models (2606.02580).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02580", "paper": "https://arxiv.org/abs/2606.02580", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.02320", "mode": "signal", "kind": "paper", "title": "TVIR: Building Deep Research Agents Towards Text--Visual Interleaved Report Generation", "summary": "Hugging Face daily paper entry for TVIR: Building Deep Research Agents Towards Text--Visual Interleaved Report Generation (2606.02320).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02320", "paper": "https://arxiv.org/abs/2606.02320", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.02255", "mode": "signal", "kind": "paper", "title": "Who Annotates in NLP? A Large-scale Assessment of Human Annotation Reporting between 2018 and 2025", "summary": "Hugging Face daily paper entry for Who Annotates in NLP? A Large-scale Assessment of Human Annotation Reporting between 2018 and 2025 (2606.02255).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02255", "paper": "https://arxiv.org/abs/2606.02255", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.02248", "mode": "signal", "kind": "paper", "title": "Geometric Latent Reasoning Induces Shorter Generations in LLMs", "summary": "Hugging Face daily paper entry for Geometric Latent Reasoning Induces Shorter Generations in LLMs (2606.02248).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.02248", "paper": "https://arxiv.org/abs/2606.02248", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01666", "mode": "signal", "kind": "paper", "title": "DOT-MoE: Differentiable Optimal Transport for MoEfication", "summary": "Hugging Face daily paper entry for DOT-MoE: Differentiable Optimal Transport for MoEfication (2606.01666).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01666", "paper": "https://arxiv.org/abs/2606.01666", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01356", "mode": "signal", "kind": "paper", "title": "A Formally Verified Library of Mathematical Finance in Lean 4", "summary": "Hugging Face daily paper entry for A Formally Verified Library of Mathematical Finance in Lean 4 (2606.01356).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01356", "paper": "https://arxiv.org/abs/2606.01356", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01348", "mode": "signal", "kind": "paper", "title": "ChartArena: Benchmarking Chart Parsing across Languages, Scenarios, and Formats", "summary": "Hugging Face daily paper entry for ChartArena: Benchmarking Chart Parsing across Languages, Scenarios, and Formats (2606.01348).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01348", "paper": "https://arxiv.org/abs/2606.01348", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01336", "mode": "signal", "kind": "paper", "title": "LongAttnComp: Cross-Family Context Compression for Long-Context Reasoning", "summary": "Hugging Face daily paper entry for LongAttnComp: Cross-Family Context Compression for Long-Context Reasoning (2606.01336).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01336", "paper": "https://arxiv.org/abs/2606.01336", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01132", "mode": "signal", "kind": "paper", "title": "HakushoBench: A Japanese Chart and Table VQA Benchmark from Governmental White Papers", "summary": "Hugging Face daily paper entry for HakushoBench: A Japanese Chart and Table VQA Benchmark from Governmental White Papers (2606.01132).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01132", "paper": "https://arxiv.org/abs/2606.01132", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01057", "mode": "signal", "kind": "paper", "title": "3DCodeBench: Benchmarking Agentic Procedural 3D Modeling Via Code", "summary": "Hugging Face daily paper entry for 3DCodeBench: Benchmarking Agentic Procedural 3D Modeling Via Code (2606.01057).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.01057", "paper": "https://arxiv.org/abs/2606.01057", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.01027", "mode": "signal", "kind": "paper", "title": "τ_0-WM: A Unified Video-Action World Model for Robotic Manipulation", "summary": "Hugging Face daily paper entry for τ_0-WM: A Unified Video-Action World Model for Robotic Manipulation (2606.01027).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2606.01027", "paper": "https://arxiv.org/abs/2606.01027", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00761", "mode": "signal", "kind": "paper", "title": "Confidence-Adaptive SwiGLU for Mixture-of-Experts", "summary": "Hugging Face daily paper entry for Confidence-Adaptive SwiGLU for Mixture-of-Experts (2606.00761).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00761", "paper": "https://arxiv.org/abs/2606.00761", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00660", "mode": "signal", "kind": "paper", "title": "FineVerify: Scaling Test-Time Compute with Fine-Grained Self-Verification for Agentic Search", "summary": "Hugging Face daily paper entry for FineVerify: Scaling Test-Time Compute with Fine-Grained Self-Verification for Agentic Search (2606.00660).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00660", "paper": "https://arxiv.org/abs/2606.00660", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00285", "mode": "signal", "kind": "paper", "title": "Model-Based Quality Assessment for Massively Multilingual Parallel Data", "summary": "Hugging Face daily paper entry for Model-Based Quality Assessment for Massively Multilingual Parallel Data (2606.00285).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00285", "paper": "https://arxiv.org/abs/2606.00285", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00267", "mode": "signal", "kind": "paper", "title": "StressDream: Steering Video World Models for Robust Policy Evaluation and Improvement", "summary": "Hugging Face daily paper entry for StressDream: Steering Video World Models for Robust Policy Evaluation and Improvement (2606.00267).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00267", "paper": "https://arxiv.org/abs/2606.00267", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00240", "mode": "signal", "kind": "paper", "title": "MindZero: Learning Online Mental Reasoning With Zero Annotations", "summary": "Hugging Face daily paper entry for MindZero: Learning Online Mental Reasoning With Zero Annotations (2606.00240).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00240", "paper": "https://arxiv.org/abs/2606.00240", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00090", "mode": "signal", "kind": "paper", "title": "Silent Failures in Physical AI: A Literature Review of Runtime Action Authorization for Autonomous Systems", "summary": "Hugging Face daily paper entry for Silent Failures in Physical AI: A Literature Review of Runtime Action Authorization for Autonomous Systems (2606.00090).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00090", "paper": "https://arxiv.org/abs/2606.00090", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2606.00089", "mode": "signal", "kind": "paper", "title": "Can Predicted Dynamics Exist in the Physical World?", "summary": "Hugging Face daily paper entry for Can Predicted Dynamics Exist in the Physical World? (2606.00089).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2606.00089", "paper": "https://arxiv.org/abs/2606.00089", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.31529", "mode": "signal", "kind": "paper", "title": "SVI-Bench: A Dynamic Microworld for Strategic Video Intelligence", "summary": "Hugging Face daily paper entry for SVI-Bench: A Dynamic Microworld for Strategic Video Intelligence (2605.31529).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.31529", "paper": "https://arxiv.org/abs/2605.31529", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.30608", "mode": "signal", "kind": "paper", "title": "Semantic Motion Anchors: Bridging Motion and Meaning in Co-Speech Gestures", "summary": "Hugging Face daily paper entry for Semantic Motion Anchors: Bridging Motion and Meaning in Co-Speech Gestures (2605.30608).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30608", "paper": "https://arxiv.org/abs/2605.30608", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.29992", "mode": "signal", "kind": "paper", "title": "Adapting Multilingual Embedding Models to Turkish via Cross-Lingual Tokenizer Surgery and Offline Distillation", "summary": "Hugging Face daily paper entry for Adapting Multilingual Embedding Models to Turkish via Cross-Lingual Tokenizer Surgery and Offline Distillation (2605.29992).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "embeddings", "tags": [ "embeddings", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29992", "paper": "https://arxiv.org/abs/2605.29992", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.29318", "mode": "signal", "kind": "paper", "title": "FreeForm: Reduced-Order Deformable Simulation from Particle-Based Skinning Eigenmodes", "summary": "Hugging Face daily paper entry for FreeForm: Reduced-Order Deformable Simulation from Particle-Based Skinning Eigenmodes (2605.29318).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29318", "paper": "https://arxiv.org/abs/2605.29318", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.29087", "mode": "signal", "kind": "paper", "title": "The Chain Holds, the Answer Folds: Trace-Answer Dissociation in Reasoning Models Under Adversarial Pressure", "summary": "Hugging Face daily paper entry for The Chain Holds, the Answer Folds: Trace-Answer Dissociation in Reasoning Models Under Adversarial Pressure (2605.29087).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29087", "paper": "https://arxiv.org/abs/2605.29087", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.29084", "mode": "signal", "kind": "paper", "title": "Same Question, Different Source, Different Answer: Auditing Source-Dependence in Medical Multi-Source RAG", "summary": "Hugging Face daily paper entry for Same Question, Different Source, Different Answer: Auditing Source-Dependence in Medical Multi-Source RAG (2605.29084).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29084", "paper": "https://arxiv.org/abs/2605.29084", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.28983", "mode": "signal", "kind": "paper", "title": "The Hamilton-Jacobi Theory of Deep Learning", "summary": "Hugging Face daily paper entry for The Hamilton-Jacobi Theory of Deep Learning (2605.28983).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28983", "paper": "https://arxiv.org/abs/2605.28983", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.28897", "mode": "signal", "kind": "paper", "title": "Review Arcade: On the Human Alignment and Gameability of LLM Reviews", "summary": "Hugging Face daily paper entry for Review Arcade: On the Human Alignment and Gameability of LLM Reviews (2605.28897).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28897", "paper": "https://arxiv.org/abs/2605.28897", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.28615", "mode": "signal", "kind": "paper", "title": "Compositional Text-to-Image Generation Via Region-aware Bimodal Direct Preference Optimization", "summary": "Hugging Face daily paper entry for Compositional Text-to-Image Generation Via Region-aware Bimodal Direct Preference Optimization (2605.28615).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28615", "paper": "https://arxiv.org/abs/2605.28615", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.28255", "mode": "signal", "kind": "paper", "title": "AI, Take the Wheel: What Drives Delegation and Trust in Human-Computer Cooperative Question Answering?", "summary": "Hugging Face daily paper entry for AI, Take the Wheel: What Drives Delegation and Trust in Human-Computer Cooperative Question Answering? (2605.28255).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28255", "paper": "https://arxiv.org/abs/2605.28255", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.27921", "mode": "signal", "kind": "paper", "title": "Show, Don't TELL: Explainable AI-Generated Text Detection", "summary": "Hugging Face daily paper entry for Show, Don't TELL: Explainable AI-Generated Text Detection (2605.27921).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27921", "paper": "https://arxiv.org/abs/2605.27921", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.26248", "mode": "signal", "kind": "paper", "title": "Unified Neural Scaling Laws", "summary": "Hugging Face daily paper entry for Unified Neural Scaling Laws (2605.26248).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26248", "paper": "https://arxiv.org/abs/2605.26248", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.25381", "mode": "signal", "kind": "paper", "title": "Not only where, But when: Temporal Scheduling for RLVR", "summary": "Hugging Face daily paper entry for Not only where, But when: Temporal Scheduling for RLVR (2605.25381).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25381", "paper": "https://arxiv.org/abs/2605.25381", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.21102", "mode": "signal", "kind": "paper", "title": "ACL-Verbatim: hallucination-free question answering for research", "summary": "Hugging Face daily paper entry for ACL-Verbatim: hallucination-free question answering for research (2605.21102).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21102", "paper": "https://arxiv.org/abs/2605.21102", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.16745", "mode": "signal", "kind": "paper", "title": "EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers", "summary": "Hugging Face daily paper entry for EVA01: Unified Native 3D Understanding and Generation via Mixture-of-Transformers (2605.16745).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16745", "paper": "https://arxiv.org/abs/2605.16745", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.09156", "mode": "signal", "kind": "paper", "title": "Lost in Translation? Exploring the Shift in Grammatical Gender from Latin to Occitan", "summary": "Hugging Face daily paper entry for Lost in Translation? Exploring the Shift in Grammatical Gender from Latin to Occitan (2605.09156).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-02", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.09156", "paper": "https://arxiv.org/abs/2605.09156", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-02", "sourceUpdatedAt": "2026-06-02", "lastSeenAt": "2026-06-03T03:04:34+00:00" }, { "id": "paper:arxiv:2605.31264", "mode": "signal", "kind": "paper", "title": "COLLEAGUE.SKILL: Automated AI Skill Generation via Expert Knowledge Distillation", "summary": "Hugging Face daily paper entry for COLLEAGUE.SKILL: Automated AI Skill Generation via Expert Knowledge Distillation (2605.31264).", "popularity": { "value": 95, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.31264", "paper": "https://arxiv.org/abs/2605.31264", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29307", "mode": "signal", "kind": "paper", "title": "GrepSeek: Training Search Agents for Direct Corpus Interaction", "summary": "Hugging Face daily paper entry for GrepSeek: Training Search Agents for Direct Corpus Interaction (2605.29307).", "popularity": { "value": 94, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29307", "paper": "https://arxiv.org/abs/2605.29307", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31159", "mode": "signal", "kind": "paper", "title": "Trust-Region Behavior Blending for On-Policy Distillation", "summary": "Hugging Face daily paper entry for Trust-Region Behavior Blending for On-Policy Distillation (2605.31159).", "popularity": { "value": 59, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31159", "paper": "https://arxiv.org/abs/2605.31159", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31604", "mode": "signal", "kind": "paper", "title": "Representation Forcing for Bottleneck-Free Unified Multimodal Models", "summary": "Hugging Face daily paper entry for Representation Forcing for Bottleneck-Free Unified Multimodal Models (2605.31604).", "popularity": { "value": 51, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31604", "paper": "https://arxiv.org/abs/2605.31604", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30993", "mode": "signal", "kind": "paper", "title": "SwanVoice: Expressive Long-Form Zero-Shot Speech Synthesis for Both Monologue and Dialogue", "summary": "Hugging Face daily paper entry for SwanVoice: Expressive Long-Form Zero-Shot Speech Synthesis for Both Monologue and Dialogue (2605.30993).", "popularity": { "value": 50, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30993", "paper": "https://arxiv.org/abs/2605.30993", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31268", "mode": "signal", "kind": "paper", "title": "Mellum2 Technical Report", "summary": "Hugging Face daily paper entry for Mellum2 Technical Report (2605.31268).", "popularity": { "value": 46, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31268", "paper": "https://arxiv.org/abs/2605.31268", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31584", "mode": "signal", "kind": "paper", "title": "LongTraceRL: Learning Long-Context Reasoning from Search Agent Trajectories with Rubric Rewards", "summary": "Hugging Face daily paper entry for LongTraceRL: Learning Long-Context Reasoning from Search Agent Trajectories with Rubric Rewards (2605.31584).", "popularity": { "value": 38, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31584", "paper": "https://arxiv.org/abs/2605.31584", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31039", "mode": "signal", "kind": "paper", "title": "GGT-100K: Generative Ground Truth for Generalizable Real-World Image Restoration", "summary": "Hugging Face daily paper entry for GGT-100K: Generative Ground Truth for Generalizable Real-World Image Restoration (2605.31039).", "popularity": { "value": 36, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.31039", "paper": "https://arxiv.org/abs/2605.31039", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30819", "mode": "signal", "kind": "paper", "title": "Function2Scene: 3D Indoor Scene Layout from Functional Specifications", "summary": "Hugging Face daily paper entry for Function2Scene: 3D Indoor Scene Layout from Functional Specifications (2605.30819).", "popularity": { "value": 36, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30819", "paper": "https://arxiv.org/abs/2605.30819", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30409", "mode": "signal", "kind": "paper", "title": "SANA-Streaming: Real-time Streaming Video Editing with Hybrid Diffusion Transformer", "summary": "Hugging Face daily paper entry for SANA-Streaming: Real-time Streaming Video Editing with Hybrid Diffusion Transformer (2605.30409).", "popularity": { "value": 32, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.30409", "paper": "https://arxiv.org/abs/2605.30409", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30940", "mode": "signal", "kind": "paper", "title": "Towards Streaming Synchronized Spatial Audio Generation via Autoregressive Diffusion Transformer", "summary": "Hugging Face daily paper entry for Towards Streaming Synchronized Spatial Audio Generation via Autoregressive Diffusion Transformer (2605.30940).", "popularity": { "value": 31, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30940", "paper": "https://arxiv.org/abs/2605.30940", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31075", "mode": "signal", "kind": "paper", "title": "Task-Focused Memorization for Multimodal Agents", "summary": "Hugging Face daily paper entry for Task-Focused Memorization for Multimodal Agents (2605.31075).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31075", "paper": "https://arxiv.org/abs/2605.31075", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30876", "mode": "signal", "kind": "paper", "title": "dMoE: dLLMs with Learnable Block Experts", "summary": "Hugging Face daily paper entry for dMoE: dLLMs with Learnable Block Experts (2605.30876).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30876", "paper": "https://arxiv.org/abs/2605.30876", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28618", "mode": "signal", "kind": "paper", "title": "Comprehensive Benchmarking of Long-Form Speech Generation in Diverse Scenarios", "summary": "Hugging Face daily paper entry for Comprehensive Benchmarking of Long-Form Speech Generation in Diverse Scenarios (2605.28618).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28618", "paper": "https://arxiv.org/abs/2605.28618", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31433", "mode": "signal", "kind": "paper", "title": "SCOPE: Self-Play via Co-Evolving Policies for Open-Ended Tasks", "summary": "Hugging Face daily paper entry for SCOPE: Self-Play via Co-Evolving Policies for Open-Ended Tasks (2605.31433).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31433", "paper": "https://arxiv.org/abs/2605.31433", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26844", "mode": "signal", "kind": "paper", "title": "Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation", "summary": "Hugging Face daily paper entry for Not All Disagreement Is Learnable: Token Teachability in On-Policy Distillation (2605.26844).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26844", "paper": "https://arxiv.org/abs/2605.26844", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29796", "mode": "signal", "kind": "paper", "title": "SAAS: Self-Aware Reinforcement Learning for Over-Search Mitigation in Agentic Search", "summary": "Hugging Face daily paper entry for SAAS: Self-Aware Reinforcement Learning for Over-Search Mitigation in Agentic Search (2605.29796).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29796", "paper": "https://arxiv.org/abs/2605.29796", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30407", "mode": "signal", "kind": "paper", "title": "Exploring Autonomous Agentic Data Engineering for Model Specialization", "summary": "Hugging Face daily paper entry for Exploring Autonomous Agentic Data Engineering for Model Specialization (2605.30407).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30407", "paper": "https://arxiv.org/abs/2605.30407", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31029", "mode": "signal", "kind": "paper", "title": "PEEK: Picking Essential frames via Efficient Knowledge distillation", "summary": "Hugging Face daily paper entry for PEEK: Picking Essential frames via Efficient Knowledge distillation (2605.31029).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.31029", "paper": "https://arxiv.org/abs/2605.31029", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30561", "mode": "signal", "kind": "paper", "title": "VLM3: Vision Language Models Are Native 3D Learners", "summary": "Hugging Face daily paper entry for VLM3: Vision Language Models Are Native 3D Learners (2605.30561).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30561", "paper": "https://arxiv.org/abs/2605.30561", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30434", "mode": "signal", "kind": "paper", "title": "LongDS-Bench: On the Failure of Long-Horizon Agentic Data Analysis", "summary": "Hugging Face daily paper entry for LongDS-Bench: On the Failure of Long-Horizon Agentic Data Analysis (2605.30434).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30434", "paper": "https://arxiv.org/abs/2605.30434", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29447", "mode": "signal", "kind": "paper", "title": "Recovering Policy-Induced Errors: Benchmarking and Trajectory Synthesis for Robust GUI Agents", "summary": "Hugging Face daily paper entry for Recovering Policy-Induced Errors: Benchmarking and Trajectory Synthesis for Robust GUI Agents (2605.29447).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29447", "paper": "https://arxiv.org/abs/2605.29447", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30621", "mode": "signal", "kind": "paper", "title": "Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents", "summary": "Hugging Face daily paper entry for Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents (2605.30621).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30621", "paper": "https://arxiv.org/abs/2605.30621", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.02772", "mode": "signal", "kind": "paper", "title": "Linearizing Vision Transformer with Test-Time Training", "summary": "Hugging Face daily paper entry for Linearizing Vision Transformer with Test-Time Training (2605.02772).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.02772", "paper": "https://arxiv.org/abs/2605.02772", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31042", "mode": "signal", "kind": "paper", "title": "From Prompt Injection to Persistent Control: Defending Agentic Harness Against Trojan Backdoors", "summary": "Hugging Face daily paper entry for From Prompt Injection to Persistent Control: Defending Agentic Harness Against Trojan Backdoors (2605.31042).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31042", "paper": "https://arxiv.org/abs/2605.31042", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31598", "mode": "signal", "kind": "paper", "title": "Linear Scaling Video VLMs for Long Video Understanding", "summary": "Hugging Face daily paper entry for Linear Scaling Video VLMs for Long Video Understanding (2605.31598).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "multimodal", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.31598", "paper": "https://arxiv.org/abs/2605.31598", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31336", "mode": "signal", "kind": "paper", "title": "DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory", "summary": "Hugging Face daily paper entry for DecMem: Towards Minute-Long Consistent World Generation with Decoupled Memory (2605.31336).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31336", "paper": "https://arxiv.org/abs/2605.31336", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28657", "mode": "signal", "kind": "paper", "title": "DEMON: Diffusion Engine for Musical Orchestrated Noise", "summary": "Hugging Face daily paper entry for DEMON: Diffusion Engine for Musical Orchestrated Noise (2605.28657).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28657", "paper": "https://arxiv.org/abs/2605.28657", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30011", "mode": "signal", "kind": "paper", "title": "VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies", "summary": "Hugging Face daily paper entry for VisualThink-VLA: Visual Intermediate Reasoning for Effective and Low-Latency Vision-Language-Action Policies (2605.30011).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30011", "paper": "https://arxiv.org/abs/2605.30011", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17543", "mode": "signal", "kind": "paper", "title": "HL-OutPaint: Coarse-to-Fine Video Outpainting for High-Resolution Long-Range Videos", "summary": "Hugging Face daily paper entry for HL-OutPaint: Coarse-to-Fine Video Outpainting for High-Resolution Long-Range Videos (2605.17543).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.17543", "paper": "https://arxiv.org/abs/2605.17543", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30846", "mode": "signal", "kind": "paper", "title": "Count Anything", "summary": "Hugging Face daily paper entry for Count Anything (2605.30846).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30846", "paper": "https://arxiv.org/abs/2605.30846", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30834", "mode": "signal", "kind": "paper", "title": "Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoring", "summary": "Hugging Face daily paper entry for Hide-and-Seek in Trajectories: Discovering Failure Signals for VLA Runtime Monitoring (2605.30834).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30834", "paper": "https://arxiv.org/abs/2605.30834", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30557", "mode": "signal", "kind": "paper", "title": "Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)?", "summary": "Hugging Face daily paper entry for Seeing Isn't Knowing: Do VLMs Know When Not to Answer Spatial Questions (and Why)? (2605.30557).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30557", "paper": "https://arxiv.org/abs/2605.30557", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31503", "mode": "signal", "kind": "paper", "title": "How can embedding models bind concepts?", "summary": "Hugging Face daily paper entry for How can embedding models bind concepts? (2605.31503).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "embeddings", "tags": [ "embeddings", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31503", "paper": "https://arxiv.org/abs/2605.31503", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30888", "mode": "signal", "kind": "paper", "title": "The Flip Side of RLHF: On-Policy Feedback for Reward Model Self-Supervised Improvement", "summary": "Hugging Face daily paper entry for The Flip Side of RLHF: On-Policy Feedback for Reward Model Self-Supervised Improvement (2605.30888).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30888", "paper": "https://arxiv.org/abs/2605.30888", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30329", "mode": "signal", "kind": "paper", "title": "SoundnessBench: Can Your AI Scientist Really Tell Good Research Ideas from Bad Ones?", "summary": "Hugging Face daily paper entry for SoundnessBench: Can Your AI Scientist Really Tell Good Research Ideas from Bad Ones? (2605.30329).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30329", "paper": "https://arxiv.org/abs/2605.30329", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31603", "mode": "signal", "kind": "paper", "title": "Lumos-Nexus: Efficient Frequency Bridging with Homogeneous Latent Space for Video Unified Models", "summary": "Hugging Face daily paper entry for Lumos-Nexus: Efficient Frequency Bridging with Homogeneous Latent Space for Video Unified Models (2605.31603).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "small-local", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.31603", "paper": "https://arxiv.org/abs/2605.31603", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31535", "mode": "signal", "kind": "paper", "title": "RayDer: Scalable Self-Supervised Novel View Synthesis from Real-World Video", "summary": "Hugging Face daily paper entry for RayDer: Scalable Self-Supervised Novel View Synthesis from Real-World Video (2605.31535).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.31535", "paper": "https://arxiv.org/abs/2605.31535", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31170", "mode": "signal", "kind": "paper", "title": "Emergent Languages in Populations of Language Model Agents: From Token Efficiency to Oversight Evasion", "summary": "Hugging Face daily paper entry for Emergent Languages in Populations of Language Model Agents: From Token Efficiency to Oversight Evasion (2605.31170).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31170", "paper": "https://arxiv.org/abs/2605.31170", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30514", "mode": "signal", "kind": "paper", "title": "MAAT: Multi-phase Adapter-Aware Targeted Unlearning", "summary": "Hugging Face daily paper entry for MAAT: Multi-phase Adapter-Aware Targeted Unlearning (2605.30514).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30514", "paper": "https://arxiv.org/abs/2605.30514", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29429", "mode": "signal", "kind": "paper", "title": "One Click per Cell Type Suffices: Training-free Group Interaction for Cell Instance Segmentation", "summary": "Hugging Face daily paper entry for One Click per Cell Type Suffices: Training-free Group Interaction for Cell Instance Segmentation (2605.29429).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.29429", "paper": "https://arxiv.org/abs/2605.29429", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28992", "mode": "signal", "kind": "paper", "title": "FRAPPE: Full Input, Residual Output Autoencoding with Projection Pursuit Encoder", "summary": "Hugging Face daily paper entry for FRAPPE: Full Input, Residual Output Autoencoding with Projection Pursuit Encoder (2605.28992).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28992", "paper": "https://arxiv.org/abs/2605.28992", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26112", "mode": "signal", "kind": "paper", "title": "From Model Scaling to System Scaling: Scaling the Harness in Agentic AI", "summary": "Hugging Face daily paper entry for From Model Scaling to System Scaling: Scaling the Harness in Agentic AI (2605.26112).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26112", "paper": "https://arxiv.org/abs/2605.26112", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23657", "mode": "signal", "kind": "paper", "title": "OpenSkillEval: Automatically Auditing the Open Skill Ecosystem for LLM Agents", "summary": "Hugging Face daily paper entry for OpenSkillEval: Automatically Auditing the Open Skill Ecosystem for LLM Agents (2605.23657).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23657", "paper": "https://arxiv.org/abs/2605.23657", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23458", "mode": "signal", "kind": "paper", "title": "One-Forcing: Towards Stable One-Step Autoregressive Video Generation", "summary": "Hugging Face daily paper entry for One-Forcing: Towards Stable One-Step Autoregressive Video Generation (2605.23458).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.23458", "paper": "https://arxiv.org/abs/2605.23458", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.31577", "mode": "signal", "kind": "paper", "title": "SurGe: Improved Surface Geometry in Point Maps", "summary": "Hugging Face daily paper entry for SurGe: Improved Surface Geometry in Point Maps (2605.31577).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.31577", "paper": "https://arxiv.org/abs/2605.31577", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.31455", "mode": "signal", "kind": "paper", "title": "DRIFT: Decoupled Rollouts and Importance-Weighted Fine-Tuning for Efficient Multi-Turn Optimization", "summary": "Hugging Face daily paper entry for DRIFT: Decoupled Rollouts and Importance-Weighted Fine-Tuning for Efficient Multi-Turn Optimization (2605.31455).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31455", "paper": "https://arxiv.org/abs/2605.31455", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.31158", "mode": "signal", "kind": "paper", "title": "Light Interaction: Training-Free Inference Acceleration for Interactive Video World Models", "summary": "Hugging Face daily paper entry for Light Interaction: Training-Free Inference Acceleration for Interactive Video World Models (2605.31158).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31158", "paper": "https://arxiv.org/abs/2605.31158", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.31096", "mode": "signal", "kind": "paper", "title": "iVGR: Internalizing Visually Grounded Reasoning for MLLMs with Reinforcement Learning", "summary": "Hugging Face daily paper entry for iVGR: Internalizing Visually Grounded Reasoning for MLLMs with Reinforcement Learning (2605.31096).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.31096", "paper": "https://arxiv.org/abs/2605.31096", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.30571", "mode": "signal", "kind": "paper", "title": "Memory-Bound but Not Bandwidth-Limited: The Physical AI Inference Gap in Batch-1 LLM Decode", "summary": "Hugging Face daily paper entry for Memory-Bound but Not Bandwidth-Limited: The Physical AI Inference Gap in Batch-1 LLM Decode (2605.30571).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30571", "paper": "https://arxiv.org/abs/2605.30571", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.29488", "mode": "signal", "kind": "paper", "title": "AnyMo: Scaling Any-Modality Conditional Motion Generation with Masked Modeling", "summary": "Hugging Face daily paper entry for AnyMo: Scaling Any-Modality Conditional Motion Generation with Masked Modeling (2605.29488).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29488", "paper": "https://arxiv.org/abs/2605.29488", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.29411", "mode": "signal", "kind": "paper", "title": "The Good, the Bad, and the Ugly of Markov Boundary for Tabular Prediction", "summary": "Hugging Face daily paper entry for The Good, the Bad, and the Ugly of Markov Boundary for Tabular Prediction (2605.29411).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29411", "paper": "https://arxiv.org/abs/2605.29411", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.29398", "mode": "signal", "kind": "paper", "title": "GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models", "summary": "Hugging Face daily paper entry for GDSD: Reinforcement Learning as Guided Denoiser Self-Distillation for Diffusion Language Models (2605.29398).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29398", "paper": "https://arxiv.org/abs/2605.29398", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.29198", "mode": "signal", "kind": "paper", "title": "Guidance Contrastive Token Credit Assignment for Discrete Policy Optimization", "summary": "Hugging Face daily paper entry for Guidance Contrastive Token Credit Assignment for Discrete Policy Optimization (2605.29198).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29198", "paper": "https://arxiv.org/abs/2605.29198", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.28969", "mode": "signal", "kind": "paper", "title": "Beyond Recall: Behavioral Specification as an Interpretive Layer for AI Personalization", "summary": "Hugging Face daily paper entry for Beyond Recall: Behavioral Specification as an Interpretive Layer for AI Personalization (2605.28969).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28969", "paper": "https://arxiv.org/abs/2605.28969", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.28730", "mode": "signal", "kind": "paper", "title": "AlphaTransit: Learning to Design City-scale Transit Routes", "summary": "Hugging Face daily paper entry for AlphaTransit: Learning to Design City-scale Transit Routes (2605.28730).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28730", "paper": "https://arxiv.org/abs/2605.28730", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.28181", "mode": "signal", "kind": "paper", "title": "When Confidence Misleads: Suffix Anchoring and Anchor-Proximity Confidence Modulation for Diffusion Language Models", "summary": "Hugging Face daily paper entry for When Confidence Misleads: Suffix Anchoring and Anchor-Proximity Confidence Modulation for Diffusion Language Models (2605.28181).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28181", "paper": "https://arxiv.org/abs/2605.28181", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.27919", "mode": "signal", "kind": "paper", "title": "Frequency-Guided Action Diffusion via Sub-Frequency Manifold Traversal", "summary": "Hugging Face daily paper entry for Frequency-Guided Action Diffusion via Sub-Frequency Manifold Traversal (2605.27919).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27919", "paper": "https://arxiv.org/abs/2605.27919", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.26562", "mode": "signal", "kind": "paper", "title": "Beyond Holistic Models: Systematic Component-level Benchmarking of Deep Multivariate Time-Series Forecasting", "summary": "Hugging Face daily paper entry for Beyond Holistic Models: Systematic Component-level Benchmarking of Deep Multivariate Time-Series Forecasting (2605.26562).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26562", "paper": "https://arxiv.org/abs/2605.26562", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.24442", "mode": "signal", "kind": "paper", "title": "Benchmarking Composed Image Retrieval for Applied Earth Observation", "summary": "Hugging Face daily paper entry for Benchmarking Composed Image Retrieval for Applied Earth Observation (2605.24442).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24442", "paper": "https://arxiv.org/abs/2605.24442", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.21625", "mode": "signal", "kind": "paper", "title": "Flat-Pack Bench: Evaluating Spatio-Temporal Understanding in Large Vision-Language Models through Furniture Assembly", "summary": "Hugging Face daily paper entry for Flat-Pack Bench: Evaluating Spatio-Temporal Understanding in Large Vision-Language Models through Furniture Assembly (2605.21625).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21625", "paper": "https://arxiv.org/abs/2605.21625", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.15779", "mode": "signal", "kind": "paper", "title": "A Topology-Aware Spatiotemporal Handover Framework for Continuous Multi-UAV Tracking", "summary": "Hugging Face daily paper entry for A Topology-Aware Spatiotemporal Handover Framework for Continuous Multi-UAV Tracking (2605.15779).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-06-01", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15779", "paper": "https://arxiv.org/abs/2605.15779", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-06-01", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-02T03:03:42+00:00" }, { "id": "paper:arxiv:2605.29801", "mode": "signal", "kind": "paper", "title": "AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security", "summary": "Hugging Face daily paper entry for AgentDoG 1.5: A Lightweight and Scalable Alignment Framework for AI Agent Safety and Security (2605.29801).", "popularity": { "value": 137, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29801", "paper": "https://arxiv.org/abs/2605.29801", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30280", "mode": "signal", "kind": "paper", "title": "Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodiments", "summary": "Hugging Face daily paper entry for Qwen-VLA: Unifying Vision-Language-Action Modeling across Tasks, Environments, and Robot Embodiments (2605.30280).", "popularity": { "value": 129, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30280", "paper": "https://arxiv.org/abs/2605.30280", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29250", "mode": "signal", "kind": "paper", "title": "OmniRetrieval: Unified Retrieval across Heterogeneous Knowledge Sources", "summary": "Hugging Face daily paper entry for OmniRetrieval: Unified Retrieval across Heterogeneous Knowledge Sources (2605.29250).", "popularity": { "value": 73, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29250", "paper": "https://arxiv.org/abs/2605.29250", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30161", "mode": "signal", "kind": "paper", "title": "Why Far Looks Up: Probing Spatial Representation in Vision-Language Models", "summary": "Hugging Face daily paper entry for Why Far Looks Up: Probing Spatial Representation in Vision-Language Models (2605.30161).", "popularity": { "value": 57, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30161", "paper": "https://arxiv.org/abs/2605.30161", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25378", "mode": "signal", "kind": "paper", "title": "CollectionLoRA: Collecting 50 Effects in 1 LoRA via Multi-Teacher On-Policy Distillation", "summary": "Hugging Face daily paper entry for CollectionLoRA: Collecting 50 Effects in 1 LoRA via Multi-Teacher On-Policy Distillation (2605.25378).", "popularity": { "value": 57, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25378", "paper": "https://arxiv.org/abs/2605.25378", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30263", "mode": "signal", "kind": "paper", "title": "minWM: A Full-Stack Open-Source Framework for Real-Time Interactive Video World Models", "summary": "Hugging Face daily paper entry for minWM: A Full-Stack Open-Source Framework for Real-Time Interactive Video World Models (2605.30263).", "popularity": { "value": 53, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.30263", "paper": "https://arxiv.org/abs/2605.30263", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30346", "mode": "signal", "kind": "paper", "title": "YoCausal: How Far is Video Generation from World Model? A Causality Perspective", "summary": "Hugging Face daily paper entry for YoCausal: How Far is Video Generation from World Model? A Causality Perspective (2605.30346).", "popularity": { "value": 46, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.30346", "paper": "https://arxiv.org/abs/2605.30346", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30260", "mode": "signal", "kind": "paper", "title": "How LoRA Remembers? A Parametric Memory Law for LLM Finetuning", "summary": "Hugging Face daily paper entry for How LoRA Remembers? A Parametric Memory Law for LLM Finetuning (2605.30260).", "popularity": { "value": 37, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30260", "paper": "https://arxiv.org/abs/2605.30260", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30248", "mode": "signal", "kind": "paper", "title": "GenClaw: Code-Driven Agentic Image Generation", "summary": "Hugging Face daily paper entry for GenClaw: Code-Driven Agentic Image Generation (2605.30248).", "popularity": { "value": 34, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30248", "paper": "https://arxiv.org/abs/2605.30248", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30010", "mode": "signal", "kind": "paper", "title": "EarlyTom: Early Token Compression Completes Fast Video Understanding", "summary": "Hugging Face daily paper entry for EarlyTom: Early Token Compression Completes Fast Video Understanding (2605.30010).", "popularity": { "value": 30, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.30010", "paper": "https://arxiv.org/abs/2605.30010", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30073", "mode": "signal", "kind": "paper", "title": "Native Audio-Visual Alignment for Generation", "summary": "Hugging Face daily paper entry for Native Audio-Visual Alignment for Generation (2605.30073).", "popularity": { "value": 28, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30073", "paper": "https://arxiv.org/abs/2605.30073", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28424", "mode": "signal", "kind": "paper", "title": "Skill0.5: Joint Skill Internalization and Utilization for Out-of-Distribution Generalization in Agentic Reinforcement Learning", "summary": "Hugging Face daily paper entry for Skill0.5: Joint Skill Internalization and Utilization for Out-of-Distribution Generalization in Agentic Reinforcement Learning (2605.28424).", "popularity": { "value": 28, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28424", "paper": "https://arxiv.org/abs/2605.28424", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29888", "mode": "signal", "kind": "paper", "title": "LaRA: Layer-wise Representation Analysis for Detecting Data Contamination in RL Post-Training", "summary": "Hugging Face daily paper entry for LaRA: Layer-wise Representation Analysis for Detecting Data Contamination in RL Post-Training (2605.29888).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29888", "paper": "https://arxiv.org/abs/2605.29888", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30076", "mode": "signal", "kind": "paper", "title": "UniSteer: Text-Guided Flow Matching in Activation Space for Versatile LLM Steering", "summary": "Hugging Face daily paper entry for UniSteer: Text-Guided Flow Matching in Activation Space for Versatile LLM Steering (2605.30076).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30076", "paper": "https://arxiv.org/abs/2605.30076", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30332", "mode": "signal", "kind": "paper", "title": "Colored Noise Diffusion Sampling", "summary": "Hugging Face daily paper entry for Colored Noise Diffusion Sampling (2605.30332).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30332", "paper": "https://arxiv.org/abs/2605.30332", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30219", "mode": "signal", "kind": "paper", "title": "When Should Models Change Their Minds? Contextual Belief Management in Large Language Models", "summary": "Hugging Face daily paper entry for When Should Models Change Their Minds? Contextual Belief Management in Large Language Models (2605.30219).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30219", "paper": "https://arxiv.org/abs/2605.30219", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30265", "mode": "signal", "kind": "paper", "title": "LoMo: Local Modality Substitution for Deeper Vision-Language Fusion", "summary": "Hugging Face daily paper entry for LoMo: Local Modality Substitution for Deeper Vision-Language Fusion (2605.30265).", "popularity": { "value": 21, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.30265", "paper": "https://arxiv.org/abs/2605.30265", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29507", "mode": "signal", "kind": "paper", "title": "Xetrieval: Mechanistically Explaining Dense Retrieval", "summary": "Hugging Face daily paper entry for Xetrieval: Mechanistically Explaining Dense Retrieval (2605.29507).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29507", "paper": "https://arxiv.org/abs/2605.29507", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26578", "mode": "signal", "kind": "paper", "title": "Is Position Bias in Dense Retrievers Built In-or Learned from Data?", "summary": "Hugging Face daily paper entry for Is Position Bias in Dense Retrievers Built In-or Learned from Data? (2605.26578).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26578", "paper": "https://arxiv.org/abs/2605.26578", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26029", "mode": "signal", "kind": "paper", "title": "CausaLab: A Scalable Environment for Interactive Causal Discovery Toward AI Scientists", "summary": "Hugging Face daily paper entry for CausaLab: A Scalable Environment for Interactive Causal Discovery Toward AI Scientists (2605.26029).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26029", "paper": "https://arxiv.org/abs/2605.26029", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29559", "mode": "signal", "kind": "paper", "title": "LiteCoder-Terminal: Scaling Long-Horizon Terminal Environments for Learning Language Agents", "summary": "Hugging Face daily paper entry for LiteCoder-Terminal: Scaling Long-Horizon Terminal Environments for Learning Language Agents (2605.29559).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29559", "paper": "https://arxiv.org/abs/2605.29559", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29341", "mode": "signal", "kind": "paper", "title": "WorldMemArena: Evaluating Multimodal Agent Memory Through Action-World Interaction", "summary": "Hugging Face daily paper entry for WorldMemArena: Evaluating Multimodal Agent Memory Through Action-World Interaction (2605.29341).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29341", "paper": "https://arxiv.org/abs/2605.29341", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27995", "mode": "signal", "kind": "paper", "title": "AsyncTool: Evaluating the Asynchronous Function Calling Capability under Multi-Task Scenarios", "summary": "Hugging Face daily paper entry for AsyncTool: Evaluating the Asynchronous Function Calling Capability under Multi-Task Scenarios (2605.27995).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27995", "paper": "https://arxiv.org/abs/2605.27995", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30102", "mode": "signal", "kind": "paper", "title": "When Cloud Agents Meet Device Agents: Lessons from Hybrid Multi-Agent Systems", "summary": "Hugging Face daily paper entry for When Cloud Agents Meet Device Agents: Lessons from Hybrid Multi-Agent Systems (2605.30102).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30102", "paper": "https://arxiv.org/abs/2605.30102", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26730", "mode": "signal", "kind": "paper", "title": "PRISM: A Multi-Dimensional Benchmark for Evaluating LLM Peer Reviewers", "summary": "Hugging Face daily paper entry for PRISM: A Multi-Dimensional Benchmark for Evaluating LLM Peer Reviewers (2605.26730).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26730", "paper": "https://arxiv.org/abs/2605.26730", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29861", "mode": "signal", "kind": "paper", "title": "Towards Verifiable Multimodal Deep Research: A Multi-Agent Harness for Interleaved Report Generation", "summary": "Hugging Face daily paper entry for Towards Verifiable Multimodal Deep Research: A Multi-Agent Harness for Interleaved Report Generation (2605.29861).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29861", "paper": "https://arxiv.org/abs/2605.29861", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29534", "mode": "signal", "kind": "paper", "title": "UI-KOBE: Knowledge-Oriented Behavior Exploration for Lightweight Graph-Guided GUI Agents", "summary": "Hugging Face daily paper entry for UI-KOBE: Knowledge-Oriented Behavior Exploration for Lightweight Graph-Guided GUI Agents (2605.29534).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29534", "paper": "https://arxiv.org/abs/2605.29534", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30349", "mode": "signal", "kind": "paper", "title": "AdaState: Self-Evolving Anchors for Streaming Video Generation", "summary": "Hugging Face daily paper entry for AdaState: Self-Evolving Anchors for Streaming Video Generation (2605.30349).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.30349", "paper": "https://arxiv.org/abs/2605.30349", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30093", "mode": "signal", "kind": "paper", "title": "Geometry Matters: 3D Foundation Priors for Learning Semantic Correspondence", "summary": "Hugging Face daily paper entry for Geometry Matters: 3D Foundation Priors for Learning Semantic Correspondence (2605.30093).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30093", "paper": "https://arxiv.org/abs/2605.30093", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29156", "mode": "signal", "kind": "paper", "title": "RUBRIC-ARROW: Alternating Pointwise Rubric Reward Modeling for LLM Post-training in Non-verifiable Domains", "summary": "Hugging Face daily paper entry for RUBRIC-ARROW: Alternating Pointwise Rubric Reward Modeling for LLM Post-training in Non-verifiable Domains (2605.29156).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29156", "paper": "https://arxiv.org/abs/2605.29156", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30350", "mode": "signal", "kind": "paper", "title": "DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation", "summary": "Hugging Face daily paper entry for DynaFLIP: Rethinking Robotics Perception via Tri-Modal-Dynamics Guided Representation (2605.30350).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30350", "paper": "https://arxiv.org/abs/2605.30350", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30347", "mode": "signal", "kind": "paper", "title": "NeuROK: Generative 4D Neural Object Kinematics", "summary": "Hugging Face daily paper entry for NeuROK: Generative 4D Neural Object Kinematics (2605.30347).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30347", "paper": "https://arxiv.org/abs/2605.30347", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30268", "mode": "signal", "kind": "paper", "title": "PhyGenHOI: Physically-Aware 4D Generation of Dynamic Human-Object Interactions", "summary": "Hugging Face daily paper entry for PhyGenHOI: Physically-Aware 4D Generation of Dynamic Human-Object Interactions (2605.30268).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30268", "paper": "https://arxiv.org/abs/2605.30268", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29157", "mode": "signal", "kind": "paper", "title": "Parallax: Parameterized Local Linear Attention for Language Modeling", "summary": "Hugging Face daily paper entry for Parallax: Parameterized Local Linear Attention for Language Modeling (2605.29157).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.29157", "paper": "https://arxiv.org/abs/2605.29157", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24785", "mode": "signal", "kind": "paper", "title": "PANDO: Efficient Multimodal AI Agents via Online Skill Distillation", "summary": "Hugging Face daily paper entry for PANDO: Efficient Multimodal AI Agents via Online Skill Distillation (2605.24785).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24785", "paper": "https://arxiv.org/abs/2605.24785", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30052", "mode": "signal", "kind": "paper", "title": "REPOT: Recoverable Program-of-Thought via Checkpoint Repair", "summary": "Hugging Face daily paper entry for REPOT: Recoverable Program-of-Thought via Checkpoint Repair (2605.30052).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30052", "paper": "https://arxiv.org/abs/2605.30052", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29648", "mode": "signal", "kind": "paper", "title": "Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering", "summary": "Hugging Face daily paper entry for Verifiable Rewards Beyond Math and Code: Lightweight Corpus-Grounded Process Supervision for Factual Question Answering (2605.29648).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.29648", "paper": "https://arxiv.org/abs/2605.29648", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29548", "mode": "signal", "kind": "paper", "title": "Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention", "summary": "Hugging Face daily paper entry for Why Larger Models Learn More: Effects of Capacity, Interference, and Rare-Task Retention (2605.29548).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29548", "paper": "https://arxiv.org/abs/2605.29548", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2601.07525", "mode": "signal", "kind": "paper", "title": "Thinking Before Constraining: A Unified Decoding Framework for Large Language Models", "summary": "Hugging Face daily paper entry for Thinking Before Constraining: A Unified Decoding Framework for Large Language Models (2601.07525).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2601.07525", "paper": "https://arxiv.org/abs/2601.07525", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29271", "mode": "signal", "kind": "paper", "title": "CoHyDE: Iterative Co-Training of LLM Rewriter & Dense Encoder for Tool Retrieval", "summary": "Hugging Face daily paper entry for CoHyDE: Iterative Co-Training of LLM Rewriter & Dense Encoder for Tool Retrieval (2605.29271).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29271", "paper": "https://arxiv.org/abs/2605.29271", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21781", "mode": "signal", "kind": "paper", "title": "Reflective Prompt Tuning through Language Model Function-Calling", "summary": "Hugging Face daily paper entry for Reflective Prompt Tuning through Language Model Function-Calling (2605.21781).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21781", "paper": "https://arxiv.org/abs/2605.21781", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.29257", "mode": "signal", "kind": "paper", "title": "ChildVox: A Speech, Audio, and Large Audio-Language Model Benchmark in Understanding and Characterizing Sound across Childhood", "summary": "Hugging Face daily paper entry for ChildVox: A Speech, Audio, and Large Audio-Language Model Benchmark in Understanding and Characterizing Sound across Childhood (2605.29257).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29257", "paper": "https://arxiv.org/abs/2605.29257", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27891", "mode": "signal", "kind": "paper", "title": "SmartDirector: Keyframe-Conditioned Cinematic Video Generation with Narrative Pacing Control", "summary": "Hugging Face daily paper entry for SmartDirector: Keyframe-Conditioned Cinematic Video Generation with Narrative Pacing Control (2605.27891).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.27891", "paper": "https://arxiv.org/abs/2605.27891", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24786", "mode": "signal", "kind": "paper", "title": "CONF-KV: Confidence-Aware KV Cache Eviction with Mixed-Precision Storage for Long-Horizon LLM", "summary": "Hugging Face daily paper entry for CONF-KV: Confidence-Aware KV Cache Eviction with Mixed-Precision Storage for Long-Horizon LLM (2605.24786).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24786", "paper": "https://arxiv.org/abs/2605.24786", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22189", "mode": "signal", "kind": "paper", "title": "Learning A Unified Risk Map for Autonomous Driving in Partially Observable Environments", "summary": "Hugging Face daily paper entry for Learning A Unified Risk Map for Autonomous Driving in Partially Observable Environments (2605.22189).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22189", "paper": "https://arxiv.org/abs/2605.22189", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30189", "mode": "signal", "kind": "paper", "title": "Token-Level Generalization in LoRA Adapter Backdoors: Attack Characterization and Behavioral Detection", "summary": "Hugging Face daily paper entry for Token-Level Generalization in LoRA Adapter Backdoors: Attack Characterization and Behavioral Detection (2605.30189).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30189", "paper": "https://arxiv.org/abs/2605.30189", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27355", "mode": "signal", "kind": "paper", "title": "Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases", "summary": "Hugging Face daily paper entry for Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases (2605.27355).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27355", "paper": "https://arxiv.org/abs/2605.27355", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25220", "mode": "signal", "kind": "paper", "title": "Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation", "summary": "Hugging Face daily paper entry for Multi-view Consistent 3D Gaussian Head Avatars 'without' Multi-view Generation (2605.25220).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25220", "paper": "https://arxiv.org/abs/2605.25220", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23235", "mode": "signal", "kind": "paper", "title": "Convex Low-resource Accent-Robust Language Detection in Speech Recognition", "summary": "Hugging Face daily paper entry for Convex Low-resource Accent-Robust Language Detection in Speech Recognition (2605.23235).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23235", "paper": "https://arxiv.org/abs/2605.23235", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.30344", "mode": "signal", "kind": "paper", "title": "Tiny but Trusted: Efficient Vision-Language Reasoning for Time-Series Anomaly Detection", "summary": "Hugging Face daily paper entry for Tiny but Trusted: Efficient Vision-Language Reasoning for Time-Series Anomaly Detection (2605.30344).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.30344", "paper": "https://arxiv.org/abs/2605.30344", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.30231", "mode": "signal", "kind": "paper", "title": "Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning", "summary": "Hugging Face daily paper entry for Beyond 3D VQAs: Injecting 3D Spatial Priors into Vision-Language Models for Enhanced Geometric Reasoning (2605.30231).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30231", "paper": "https://arxiv.org/abs/2605.30231", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.30060", "mode": "signal", "kind": "paper", "title": "Towards Consistent Video Geometry Estimation", "summary": "Hugging Face daily paper entry for Towards Consistent Video Geometry Estimation (2605.30060).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.30060", "paper": "https://arxiv.org/abs/2605.30060", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.30003", "mode": "signal", "kind": "paper", "title": "Discovering Cooperative Pipelines: Autoresearch for Sequential Social Dilemmas", "summary": "Hugging Face daily paper entry for Discovering Cooperative Pipelines: Autoresearch for Sequential Social Dilemmas (2605.30003).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.30003", "paper": "https://arxiv.org/abs/2605.30003", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.29486", "mode": "signal", "kind": "paper", "title": "PhoneWorld: Scaling Phone-Use Agent Environments", "summary": "Hugging Face daily paper entry for PhoneWorld: Scaling Phone-Use Agent Environments (2605.29486).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.29486", "paper": "https://arxiv.org/abs/2605.29486", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.26485", "mode": "signal", "kind": "paper", "title": "OmniInteract: Benchmarking Real-World Streaming Interaction for Real-Time Omnimodal Assistants", "summary": "Hugging Face daily paper entry for OmniInteract: Benchmarking Real-World Streaming Interaction for Real-Time Omnimodal Assistants (2605.26485).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26485", "paper": "https://arxiv.org/abs/2605.26485", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22771", "mode": "signal", "kind": "paper", "title": "Reducing Political Manipulation with Consistency Training", "summary": "Hugging Face daily paper entry for Reducing Political Manipulation with Consistency Training (2605.22771).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22771", "paper": "https://arxiv.org/abs/2605.22771", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22765", "mode": "signal", "kind": "paper", "title": "Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation", "summary": "Hugging Face daily paper entry for Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation (2605.22765).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22765", "paper": "https://arxiv.org/abs/2605.22765", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.16363", "mode": "signal", "kind": "paper", "title": "ORACLE: Anticipating Scams from Partial Trajectories in Streaming App Usage", "summary": "Hugging Face daily paper entry for ORACLE: Anticipating Scams from Partial Trajectories in Streaming App Usage (2605.16363).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16363", "paper": "https://arxiv.org/abs/2605.16363", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.13857", "mode": "signal", "kind": "paper", "title": "MoZoo:Unleashing Video Diffusion power in animal fur and muscle simulation", "summary": "Hugging Face daily paper entry for MoZoo:Unleashing Video Diffusion power in animal fur and muscle simulation (2605.13857).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-31", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.13857", "paper": "https://arxiv.org/abs/2605.13857", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-31", "sourceUpdatedAt": "2026-05-31", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.28816", "mode": "signal", "kind": "paper", "title": "Gamma-World: Generative Multi-Agent World Modeling Beyond Two Players", "summary": "A generative multi-agent world model is presented that uses simplex rotary agent encoding and sparse hub attention to enable scalable, permutation-symmetric interaction between multiple agents in interactive video generation. World models for interactive vi...", "popularity": { "value": 417, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28816", "paper": "https://arxiv.org/abs/2605.28816", "repo": "", "docs": "https://arxiv.org/pdf/2605.28816" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28774", "mode": "signal", "kind": "paper", "title": "Agent Explorative Policy Optimization for Multimodal Agentic Reasoning", "summary": "Hugging Face daily paper entry for Agent Explorative Policy Optimization for Multimodal Agentic Reasoning (2605.28774).", "popularity": { "value": 86, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28774", "paper": "https://arxiv.org/abs/2605.28774", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28293", "mode": "signal", "kind": "paper", "title": "ProRL: Effective Reinforcement Learning for Proactive Recommendation via Rectified Policy Gradient Estimation", "summary": "Hugging Face daily paper entry for ProRL: Effective Reinforcement Learning for Proactive Recommendation via Rectified Policy Gradient Estimation (2605.28293).", "popularity": { "value": 85, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28293", "paper": "https://arxiv.org/abs/2605.28293", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28820", "mode": "signal", "kind": "paper", "title": "From Pixels to Words -- Towards Native One-Vision Models at Scale", "summary": "Hugging Face daily paper entry for From Pixels to Words -- Towards Native One-Vision Models at Scale (2605.28820).", "popularity": { "value": 70, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.28820", "paper": "https://arxiv.org/abs/2605.28820", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28814", "mode": "signal", "kind": "paper", "title": "Self-Improving Language Models with Bidirectional Evolutionary Search", "summary": "Hugging Face daily paper entry for Self-Improving Language Models with Bidirectional Evolutionary Search (2605.28814).", "popularity": { "value": 56, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28814", "paper": "https://arxiv.org/abs/2605.28814", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28003", "mode": "signal", "kind": "paper", "title": "ResearchMath-14K: Scaling Research-Level Mathematics via Agents", "summary": "Hugging Face daily paper entry for ResearchMath-14K: Scaling Research-Level Mathematics via Agents (2605.28003).", "popularity": { "value": 49, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28003", "paper": "https://arxiv.org/abs/2605.28003", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28421", "mode": "signal", "kind": "paper", "title": "DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes", "summary": "Hugging Face daily paper entry for DenoiseRL: Bootstrapping Reasoning Models to Recover from Noisy Prefixes (2605.28421).", "popularity": { "value": 46, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28421", "paper": "https://arxiv.org/abs/2605.28421", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28548", "mode": "signal", "kind": "paper", "title": "GEM: Generative Supervision Helps Embodied Intelligence", "summary": "Hugging Face daily paper entry for GEM: Generative Supervision Helps Embodied Intelligence (2605.28548).", "popularity": { "value": 40, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.28548", "paper": "https://arxiv.org/abs/2605.28548", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28732", "mode": "signal", "kind": "paper", "title": "MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems", "summary": "Hugging Face daily paper entry for MemTrace: Tracing and Attributing Errors in Large Language Model Memory Systems (2605.28732).", "popularity": { "value": 39, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28732", "paper": "https://arxiv.org/abs/2605.28732", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28775", "mode": "signal", "kind": "paper", "title": "Learn from Weaknesses: Automated Domain Specialization for Small Computer-Use Agents", "summary": "Hugging Face daily paper entry for Learn from Weaknesses: Automated Domain Specialization for Small Computer-Use Agents (2605.28775).", "popularity": { "value": 37, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28775", "paper": "https://arxiv.org/abs/2605.28775", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26340", "mode": "signal", "kind": "paper", "title": "ScientistOne: Towards Human-Level Autonomous Research via Chain-of-Evidence", "summary": "Hugging Face daily paper entry for ScientistOne: Towards Human-Level Autonomous Research via Chain-of-Evidence (2605.26340).", "popularity": { "value": 35, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26340", "paper": "https://arxiv.org/abs/2605.26340", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28773", "mode": "signal", "kind": "paper", "title": "Rethinking Memory as Continuously Evolving Connectivity", "summary": "Hugging Face daily paper entry for Rethinking Memory as Continuously Evolving Connectivity (2605.28773).", "popularity": { "value": 32, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28773", "paper": "https://arxiv.org/abs/2605.28773", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26302", "mode": "signal", "kind": "paper", "title": "Your Agents Are Aging Too: Agent Lifespan Engineering for Deployed Systems", "summary": "Hugging Face daily paper entry for Your Agents Are Aging Too: Agent Lifespan Engineering for Deployed Systems (2605.26302).", "popularity": { "value": 31, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26302", "paper": "https://arxiv.org/abs/2605.26302", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27760", "mode": "signal", "kind": "paper", "title": "SkillGrad: Optimizing Agent Skills Like Gradient Descent", "summary": "Hugging Face daily paper entry for SkillGrad: Optimizing Agent Skills Like Gradient Descent (2605.27760).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27760", "paper": "https://arxiv.org/abs/2605.27760", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27905", "mode": "signal", "kind": "paper", "title": "AI Research Agents Narrow Scientific Exploration", "summary": "Hugging Face daily paper entry for AI Research Agents Narrow Scientific Exploration (2605.27905).", "popularity": { "value": 25, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27905", "paper": "https://arxiv.org/abs/2605.27905", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28691", "mode": "signal", "kind": "paper", "title": "OSP-Next: Efficient High-Quality Video Generation with Sparse Sequence Parallelism, HiF8 Quantization, and Reinforcement Learning", "summary": "Hugging Face daily paper entry for OSP-Next: Efficient High-Quality Video Generation with Sparse Sequence Parallelism, HiF8 Quantization, and Reinforcement Learning (2605.28691).", "popularity": { "value": 24, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28691", "paper": "https://arxiv.org/abs/2605.28691", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28534", "mode": "signal", "kind": "paper", "title": "GUI-CIDER: Mid-training GUI Agents via Causal Internalization and Density-aware Exemplar Reselection", "summary": "Hugging Face daily paper entry for GUI-CIDER: Mid-training GUI Agents via Causal Internalization and Density-aware Exemplar Reselection (2605.28534).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28534", "paper": "https://arxiv.org/abs/2605.28534", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28109", "mode": "signal", "kind": "paper", "title": "Long Live The Balance: Information Bottleneck Driven Tree-based Policy Optimization", "summary": "Hugging Face daily paper entry for Long Live The Balance: Information Bottleneck Driven Tree-based Policy Optimization (2605.28109).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28109", "paper": "https://arxiv.org/abs/2605.28109", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25969", "mode": "signal", "kind": "paper", "title": "Triplet-Block Diffusion RWKV", "summary": "Hugging Face daily paper entry for Triplet-Block Diffusion RWKV (2605.25969).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25969", "paper": "https://arxiv.org/abs/2605.25969", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27310", "mode": "signal", "kind": "paper", "title": "How and What to Imagine? Visual Thinking in Unified Multimodal Models for Cross-View Spatial Reasoning", "summary": "Hugging Face daily paper entry for How and What to Imagine? Visual Thinking in Unified Multimodal Models for Cross-View Spatial Reasoning (2605.27310).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27310", "paper": "https://arxiv.org/abs/2605.27310", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26396", "mode": "signal", "kind": "paper", "title": "Advancing Creative Physical Intelligence in Large Multimodal Models", "summary": "Hugging Face daily paper entry for Advancing Creative Physical Intelligence in Large Multimodal Models (2605.26396).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26396", "paper": "https://arxiv.org/abs/2605.26396", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27491", "mode": "signal", "kind": "paper", "title": "GE-Sim 2.0: A Roadmap Towards Comprehensive Closed-loop Video World Simulators for Robotic Manipulation", "summary": "Hugging Face daily paper entry for GE-Sim 2.0: A Roadmap Towards Comprehensive Closed-loop Video World Simulators for Robotic Manipulation (2605.27491).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.27491", "paper": "https://arxiv.org/abs/2605.27491", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23163", "mode": "signal", "kind": "paper", "title": "Fast-dDrive: Efficient Block-Diffusion VLM for Autonomous Driving", "summary": "Hugging Face daily paper entry for Fast-dDrive: Efficient Block-Diffusion VLM for Autonomous Driving (2605.23163).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.23163", "paper": "https://arxiv.org/abs/2605.23163", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28721", "mode": "signal", "kind": "paper", "title": "LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know?", "summary": "Hugging Face daily paper entry for LiveBrowseComp: Are Search Agents Searching, or Just Verifying What They Already Know? (2605.28721).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28721", "paper": "https://arxiv.org/abs/2605.28721", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28398", "mode": "signal", "kind": "paper", "title": "HRBench: Benchmarking and Understanding Thinking-Mode Switch Strategies in Hybrid-Reasoning LLMs", "summary": "Hugging Face daily paper entry for HRBench: Benchmarking and Understanding Thinking-Mode Switch Strategies in Hybrid-Reasoning LLMs (2605.28398).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28398", "paper": "https://arxiv.org/abs/2605.28398", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27354", "mode": "signal", "kind": "paper", "title": "Guiding LLM Post-training Data Engineering with Model Internals from Sparse Autoencoders", "summary": "Hugging Face daily paper entry for Guiding LLM Post-training Data Engineering with Model Internals from Sparse Autoencoders (2605.27354).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27354", "paper": "https://arxiv.org/abs/2605.27354", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28763", "mode": "signal", "kind": "paper", "title": "CubePart: An Open-Vocabulary Part-Controllable 3D Generator", "summary": "Hugging Face daily paper entry for CubePart: An Open-Vocabulary Part-Controllable 3D Generator (2605.28763).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28763", "paper": "https://arxiv.org/abs/2605.28763", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27882", "mode": "signal", "kind": "paper", "title": "VibeSearchBench: Benchmarking Long-horizon Proactive Search in the Wild", "summary": "Hugging Face daily paper entry for VibeSearchBench: Benchmarking Long-horizon Proactive Search in the Wild (2605.27882).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27882", "paper": "https://arxiv.org/abs/2605.27882", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27268", "mode": "signal", "kind": "paper", "title": "Lost in Sampling: Assessing Lexical Reachability in LLMs via the Word Coverage Score (WCS)", "summary": "Hugging Face daily paper entry for Lost in Sampling: Assessing Lexical Reachability in LLMs via the Word Coverage Score (WCS) (2605.27268).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27268", "paper": "https://arxiv.org/abs/2605.27268", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27028", "mode": "signal", "kind": "paper", "title": "Less is More: Early Stopping Rollout for On-Policy Distillation", "summary": "Hugging Face daily paper entry for Less is More: Early Stopping Rollout for On-Policy Distillation (2605.27028).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27028", "paper": "https://arxiv.org/abs/2605.27028", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26032", "mode": "signal", "kind": "paper", "title": "Everything at Every Scale: Scale-Invariant Diffusion with Continuous Super-Resolution", "summary": "Hugging Face daily paper entry for Everything at Every Scale: Scale-Invariant Diffusion with Continuous Super-Resolution (2605.26032).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26032", "paper": "https://arxiv.org/abs/2605.26032", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27901", "mode": "signal", "kind": "paper", "title": "The Fragility of Chain-of-Thought Monitoring Across Typologically Diverse Languages", "summary": "Hugging Face daily paper entry for The Fragility of Chain-of-Thought Monitoring Across Typologically Diverse Languages (2605.27901).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27901", "paper": "https://arxiv.org/abs/2605.27901", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26574", "mode": "signal", "kind": "paper", "title": "GradSentry: Gradient Spectral Entropy for Backdoor Sample Filtering in Large Language Model Fine-Tuning", "summary": "Hugging Face daily paper entry for GradSentry: Gradient Spectral Entropy for Backdoor Sample Filtering in Large Language Model Fine-Tuning (2605.26574).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26574", "paper": "https://arxiv.org/abs/2605.26574", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28805", "mode": "signal", "kind": "paper", "title": "OmniVerifier-M1: Multimodal Meta-Verifier with Explicit Structured Recalibration", "summary": "Hugging Face daily paper entry for OmniVerifier-M1: Multimodal Meta-Verifier with Explicit Structured Recalibration (2605.28805).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28805", "paper": "https://arxiv.org/abs/2605.28805", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28655", "mode": "signal", "kind": "paper", "title": "AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation", "summary": "Hugging Face daily paper entry for AutoScientists: Self-Organizing Agent Teams for Long-Running Scientific Experimentation (2605.28655).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28655", "paper": "https://arxiv.org/abs/2605.28655", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19952", "mode": "signal", "kind": "paper", "title": "Rethinking How to Remember: Beyond Atomic Facts in Lifelong LLM Agent Memory", "summary": "Hugging Face daily paper entry for Rethinking How to Remember: Beyond Atomic Facts in Lifelong LLM Agent Memory (2605.19952).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19952", "paper": "https://arxiv.org/abs/2605.19952", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28819", "mode": "signal", "kind": "paper", "title": "PEFT-Arena: Understanding Parameter-Efficient Finetuning from a Stability-Plasticity Perspective", "summary": "Hugging Face daily paper entry for PEFT-Arena: Understanding Parameter-Efficient Finetuning from a Stability-Plasticity Perspective (2605.28819).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28819", "paper": "https://arxiv.org/abs/2605.28819", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28591", "mode": "signal", "kind": "paper", "title": "Models That Know How Evaluations Are Designed Score Safer", "summary": "Hugging Face daily paper entry for Models That Know How Evaluations Are Designed Score Safer (2605.28591).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28591", "paper": "https://arxiv.org/abs/2605.28591", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28257", "mode": "signal", "kind": "paper", "title": "Category-Level 3D Correspondence in Camera Space via Morphable Object Priors", "summary": "Hugging Face daily paper entry for Category-Level 3D Correspondence in Camera Space via Morphable Object Priors (2605.28257).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28257", "paper": "https://arxiv.org/abs/2605.28257", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27466", "mode": "signal", "kind": "paper", "title": "AgensFlow: A Coordination-Policy Substrate for Multi-Agent Systems", "summary": "Hugging Face daily paper entry for AgensFlow: A Coordination-Policy Substrate for Multi-Agent Systems (2605.27466).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27466", "paper": "https://arxiv.org/abs/2605.27466", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28617", "mode": "signal", "kind": "paper", "title": "LACUNA: Safe Agents as Recursive Program Holes", "summary": "Hugging Face daily paper entry for LACUNA: Safe Agents as Recursive Program Holes (2605.28617).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28617", "paper": "https://arxiv.org/abs/2605.28617", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27762", "mode": "signal", "kind": "paper", "title": "PEAM: Parametric Embodied Agent Memory through Contrastive Internalization of Experience in Minecraft", "summary": "Hugging Face daily paper entry for PEAM: Parametric Embodied Agent Memory through Contrastive Internalization of Experience in Minecraft (2605.27762).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27762", "paper": "https://arxiv.org/abs/2605.27762", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28184", "mode": "signal", "kind": "paper", "title": "Joint Training of Multi-Token Prediction in Reinforcement Learning via Optimal Coefficient Calibration", "summary": "Hugging Face daily paper entry for Joint Training of Multi-Token Prediction in Reinforcement Learning via Optimal Coefficient Calibration (2605.28184).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.28184", "paper": "https://arxiv.org/abs/2605.28184", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28158", "mode": "signal", "kind": "paper", "title": "OR-Space: A Full-Lifecycle Workspace Benchmark for Industrial Optimization Agents", "summary": "Hugging Face daily paper entry for OR-Space: A Full-Lifecycle Workspace Benchmark for Industrial Optimization Agents (2605.28158).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28158", "paper": "https://arxiv.org/abs/2605.28158", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27908", "mode": "signal", "kind": "paper", "title": "ESC-Skills: Discovering and Self-Evolving Skills for Emotional Support Conversations", "summary": "Hugging Face daily paper entry for ESC-Skills: Discovering and Self-Evolving Skills for Emotional Support Conversations (2605.27908).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27908", "paper": "https://arxiv.org/abs/2605.27908", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26457", "mode": "signal", "kind": "paper", "title": "Verus-SpecGym: An Agentic Environment for Evaluating Specification Autoformalization", "summary": "Hugging Face daily paper entry for Verus-SpecGym: An Agentic Environment for Evaluating Specification Autoformalization (2605.26457).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26457", "paper": "https://arxiv.org/abs/2605.26457", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25707", "mode": "signal", "kind": "paper", "title": "AgentHijack: Benchmarking Computer Use Agent Robustness to Common Environment Corruptions", "summary": "Hugging Face daily paper entry for AgentHijack: Benchmarking Computer Use Agent Robustness to Common Environment Corruptions (2605.25707).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25707", "paper": "https://arxiv.org/abs/2605.25707", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24486", "mode": "signal", "kind": "paper", "title": "AgentFugue: Agent Scaling for Long-Horizon Tasks through Collective Reasoning", "summary": "Hugging Face daily paper entry for AgentFugue: Agent Scaling for Long-Horizon Tasks through Collective Reasoning (2605.24486).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24486", "paper": "https://arxiv.org/abs/2605.24486", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28510", "mode": "signal", "kind": "paper", "title": "Efficient and Scalable Provenance Tracking for LLM-Generated Code Snippets", "summary": "Hugging Face daily paper entry for Efficient and Scalable Provenance Tracking for LLM-Generated Code Snippets (2605.28510).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.28510", "paper": "https://arxiv.org/abs/2605.28510", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24053", "mode": "signal", "kind": "paper", "title": "Breaking the Chains of Probability: Neutrosophic Logic as a New Framework for Epistemic Uncertainty in Large Language Models", "summary": "Hugging Face daily paper entry for Breaking the Chains of Probability: Neutrosophic Logic as a New Framework for Epistemic Uncertainty in Large Language Models (2605.24053).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24053", "paper": "https://arxiv.org/abs/2605.24053", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.28034", "mode": "signal", "kind": "paper", "title": "Clark Hash: Stateless Sparse Johnson-Lindenstrauss Quantization for Neural Embeddings", "summary": "Hugging Face daily paper entry for Clark Hash: Stateless Sparse Johnson-Lindenstrauss Quantization for Neural Embeddings (2605.28034).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.28034", "paper": "https://arxiv.org/abs/2605.28034", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.27824", "mode": "signal", "kind": "paper", "title": "Revealing Algorithmic Deductive Circuits for Logical Reasoning", "summary": "Hugging Face daily paper entry for Revealing Algorithmic Deductive Circuits for Logical Reasoning (2605.27824).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27824", "paper": "https://arxiv.org/abs/2605.27824", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.27766", "mode": "signal", "kind": "paper", "title": "Got a Secret? LLM Agents Can't Keep It: Evaluating Privacy in Multi-Agent Systems", "summary": "Hugging Face daily paper entry for Got a Secret? LLM Agents Can't Keep It: Evaluating Privacy in Multi-Agent Systems (2605.27766).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27766", "paper": "https://arxiv.org/abs/2605.27766", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.27311", "mode": "signal", "kind": "paper", "title": "Chartographer: Counterfactual Chart Generation for Evaluating Vision-Language Models", "summary": "Hugging Face daily paper entry for Chartographer: Counterfactual Chart Generation for Evaluating Vision-Language Models (2605.27311).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27311", "paper": "https://arxiv.org/abs/2605.27311", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.27044", "mode": "signal", "kind": "paper", "title": "BatteryMFormer: Multi-level Learning for Battery Degradation Trajectory Forecasting", "summary": "Hugging Face daily paper entry for BatteryMFormer: Multi-level Learning for Battery Degradation Trajectory Forecasting (2605.27044).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27044", "paper": "https://arxiv.org/abs/2605.27044", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.26368", "mode": "signal", "kind": "paper", "title": "Unified Panoramic Geometry Estimation via Multi-View Foundation Models", "summary": "Hugging Face daily paper entry for Unified Panoramic Geometry Estimation via Multi-View Foundation Models (2605.26368).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26368", "paper": "https://arxiv.org/abs/2605.26368", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.25940", "mode": "signal", "kind": "paper", "title": "How Accurate are Video Quality Models for Diffusion-Based Video Super-Resolution?", "summary": "Hugging Face daily paper entry for How Accurate are Video Quality Models for Diffusion-Based Video Super-Resolution? (2605.25940).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.25940", "paper": "https://arxiv.org/abs/2605.25940", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.25174", "mode": "signal", "kind": "paper", "title": "Growing a Neural Network in Breadth, Depth, and Time", "summary": "Hugging Face daily paper entry for Growing a Neural Network in Breadth, Depth, and Time (2605.25174).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25174", "paper": "https://arxiv.org/abs/2605.25174", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.23346", "mode": "signal", "kind": "paper", "title": "Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion", "summary": "Hugging Face daily paper entry for Contrastive Distribution Matching for Amortized Sequential Monte Carlo in Discrete Diffusion (2605.23346).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23346", "paper": "https://arxiv.org/abs/2605.23346", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17531", "mode": "signal", "kind": "paper", "title": "Don't Guess, Just Ask: Resolving Ambiguity in Referring Segmentation via Multi-turn Clarification", "summary": "Hugging Face daily paper entry for Don't Guess, Just Ask: Resolving Ambiguity in Referring Segmentation via Multi-turn Clarification (2605.17531).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-28", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.17531", "paper": "https://arxiv.org/abs/2605.17531", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-28", "sourceUpdatedAt": "2026-05-28", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "github:helloianneo/ian-xiaohei-illustrations", "mode": "catalog", "kind": "repo", "title": "helloianneo/ian-xiaohei-illustrations", "summary": "中文小黑怪诞正文配图生成 Skill | 16:9 白底手绘 | 少量红橙蓝批注 | Codex Skill", "popularity": { "value": 1973, "label": "stars" }, "whyItMatters": "Useful to archive because it is a source-listed GitHub repository with measurable stars.", "sourceId": "github-catalog", "sourceName": "GitHub Catalog", "sourceUrl": "https://api.github.com/search/repositories?q=topic:llm&sort=stars&order=desc&per_page=2", "publishedAt": "2026-06-03", "category": "frameworks", "subcategory": "agents", "tags": [ "agents", "ai-agent" ], "links": { "primary": "https://github.com/helloianneo/ian-xiaohei-illustrations", "paper": "", "repo": "https://github.com/helloianneo/ian-xiaohei-illustrations", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": true, "license": "MIT", "status": "active", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-06-03", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "github:op7418/guizang-social-card-skill", "mode": "catalog", "kind": "repo", "title": "op7418/guizang-social-card-skill", "summary": "🪧 Claude Code / Codex skill — generate Xiaohongshu carousels & WeChat 21:9+1:1 cover pairs. Editorial × Swiss visual systems, 28 layouts, 10 themes, single-file HTML → PNG. 小红书图文 + 公众号封面对", "popularity": { "value": 2748, "label": "stars" }, "whyItMatters": "Useful to archive because it is a source-listed GitHub repository with measurable stars.", "sourceId": "github-catalog", "sourceName": "GitHub Catalog", "sourceUrl": "https://api.github.com/search/repositories?q=topic:llm&sort=stars&order=desc&per_page=2", "publishedAt": "2026-05-27", "category": "frameworks", "subcategory": "agents", "tags": [ "agents", "ai-agent" ], "links": { "primary": "https://github.com/op7418/guizang-social-card-skill", "paper": "", "repo": "https://github.com/op7418/guizang-social-card-skill", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": true, "license": "AGPL-3.0", "status": "active", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.27365", "mode": "signal", "kind": "paper", "title": "LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding", "summary": "Hugging Face daily paper entry for LocateAnything: Fast and High-Quality Vision-Language Grounding with Parallel Box Decoding (2605.27365).", "popularity": { "value": 134, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27365", "paper": "https://arxiv.org/abs/2605.27365", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23271", "mode": "signal", "kind": "paper", "title": "EvalVerse: Pipeline-Aware and Expert-Calibrated Benchmarking for Professional Cinematic Video Generation", "summary": "Hugging Face daily paper entry for EvalVerse: Pipeline-Aware and Expert-Calibrated Benchmarking for Professional Cinematic Video Generation (2605.23271).", "popularity": { "value": 79, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23271", "paper": "https://arxiv.org/abs/2605.23271", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27367", "mode": "signal", "kind": "paper", "title": "SpatialBench: Is Your Spatial Foundation Model an All-Round Player?", "summary": "Hugging Face daily paper entry for SpatialBench: Is Your Spatial Foundation Model an All-Round Player? (2605.27367).", "popularity": { "value": 70, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27367", "paper": "https://arxiv.org/abs/2605.27367", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26114", "mode": "signal", "kind": "paper", "title": "MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent Research", "summary": "Hugging Face daily paper entry for MobileGym: A Verifiable and Highly Parallel Simulation Platform for Mobile GUI Agent Research (2605.26114).", "popularity": { "value": 62, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26114", "paper": "https://arxiv.org/abs/2605.26114", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26230", "mode": "signal", "kind": "paper", "title": "Geometry-Aware Representation Denoising for Robust Multi-view 3D Reconstruction", "summary": "Hugging Face daily paper entry for Geometry-Aware Representation Denoising for Robust Multi-view 3D Reconstruction (2605.26230).", "popularity": { "value": 41, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26230", "paper": "https://arxiv.org/abs/2605.26230", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26494", "mode": "signal", "kind": "paper", "title": "The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence", "summary": "Hugging Face daily paper entry for The MiniMax-M2 Series: Mini Activations Unleashing Max Real-World Intelligence (2605.26494).", "popularity": { "value": 39, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26494", "paper": "https://arxiv.org/abs/2605.26494", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26244", "mode": "signal", "kind": "paper", "title": "LongAV-Compass: Towards Unified Evaluation of Minute-Scale Audio-Visual Generation Across T2AV, I2AV, and V2AV", "summary": "Hugging Face daily paper entry for LongAV-Compass: Towards Unified Evaluation of Minute-Scale Audio-Visual Generation Across T2AV, I2AV, and V2AV (2605.26244).", "popularity": { "value": 38, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26244", "paper": "https://arxiv.org/abs/2605.26244", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25893", "mode": "signal", "kind": "paper", "title": "D^2-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation-Aware Routing", "summary": "Hugging Face daily paper entry for D^2-Monitor: Dynamic Safety Monitoring for Diffusion LLMs via Hesitation-Aware Routing (2605.25893).", "popularity": { "value": 38, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25893", "paper": "https://arxiv.org/abs/2605.25893", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17423", "mode": "signal", "kind": "paper", "title": "Soap2Soap: Long Cinematic Video Remaking via Multi-Agent Collaboration", "summary": "Hugging Face daily paper entry for Soap2Soap: Long Cinematic Video Remaking via Multi-Agent Collaboration (2605.17423).", "popularity": { "value": 33, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17423", "paper": "https://arxiv.org/abs/2605.17423", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27102", "mode": "signal", "kind": "paper", "title": "JLT: Clean-Latent Prediction in Latent Diffusion Transformers", "summary": "Hugging Face daily paper entry for JLT: Clean-Latent Prediction in Latent Diffusion Transformers (2605.27102).", "popularity": { "value": 31, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27102", "paper": "https://arxiv.org/abs/2605.27102", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27030", "mode": "signal", "kind": "paper", "title": "Share More, Search Less: Collaborative Parallel Thinking for Efficient Test-Time Scaling", "summary": "Hugging Face daily paper entry for Share More, Search Less: Collaborative Parallel Thinking for Efficient Test-Time Scaling (2605.27030).", "popularity": { "value": 31, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.27030", "paper": "https://arxiv.org/abs/2605.27030", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25979", "mode": "signal", "kind": "paper", "title": "LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence", "summary": "Hugging Face daily paper entry for LLaVA-OneVision-2: Towards Next-Generation Perceptual Intelligence (2605.25979).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.25979", "paper": "https://arxiv.org/abs/2605.25979", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27366", "mode": "signal", "kind": "paper", "title": "MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation", "summary": "Hugging Face daily paper entry for MUSE-Autoskill: Self-Evolving Agents via Skill Creation, Memory, Management, and Evaluation (2605.27366).", "popularity": { "value": 25, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27366", "paper": "https://arxiv.org/abs/2605.27366", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27295", "mode": "signal", "kind": "paper", "title": "Gemini Embedding 2: A Native Multimodal Embedding Model from Gemini", "summary": "Hugging Face daily paper entry for Gemini Embedding 2: A Native Multimodal Embedding Model from Gemini (2605.27295).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "embeddings", "tags": [ "embeddings", "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27295", "paper": "https://arxiv.org/abs/2605.27295", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27068", "mode": "signal", "kind": "paper", "title": "QUACK: Questioning, Understanding, and Auditing Communicated Knowledge in Multimodal Social Deduction Agents", "summary": "Hugging Face daily paper entry for QUACK: Questioning, Understanding, and Auditing Communicated Knowledge in Multimodal Social Deduction Agents (2605.27068).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27068", "paper": "https://arxiv.org/abs/2605.27068", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27476", "mode": "signal", "kind": "paper", "title": "Balancing Fidelity and Diversity in Diffusion Models via Symmetric Attention Decomposition: Hopfield Perspective", "summary": "Hugging Face daily paper entry for Balancing Fidelity and Diversity in Diffusion Models via Symmetric Attention Decomposition: Hopfield Perspective (2605.27476).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27476", "paper": "https://arxiv.org/abs/2605.27476", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26895", "mode": "signal", "kind": "paper", "title": "Negligible in Size, Significant in Effect: On Scale Vectors in Large Language Models", "summary": "Hugging Face daily paper entry for Negligible in Size, Significant in Effect: On Scale Vectors in Large Language Models (2605.26895).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26895", "paper": "https://arxiv.org/abs/2605.26895", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27141", "mode": "signal", "kind": "paper", "title": "VitaBench 2.0: Evaluating Personalized and Proactive Agents in Long-Term User Interactions", "summary": "Hugging Face daily paper entry for VitaBench 2.0: Evaluating Personalized and Proactive Agents in Long-Term User Interactions (2605.27141).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27141", "paper": "https://arxiv.org/abs/2605.27141", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27209", "mode": "signal", "kind": "paper", "title": "Learning to Act under Noise: Enhancing Agent Robustness via Noisy Environments", "summary": "Hugging Face daily paper entry for Learning to Act under Noise: Enhancing Agent Robustness via Noisy Environments (2605.27209).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.27209", "paper": "https://arxiv.org/abs/2605.27209", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26952", "mode": "signal", "kind": "paper", "title": "Efficient Agentic Reinforcement Learning with On-Policy Intrinsic Knowledge Boundary Enhancement", "summary": "Hugging Face daily paper entry for Efficient Agentic Reinforcement Learning with On-Policy Intrinsic Knowledge Boundary Enhancement (2605.26952).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26952", "paper": "https://arxiv.org/abs/2605.26952", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25437", "mode": "signal", "kind": "paper", "title": "Does Seeing More Mean Knowing More? Mono-Anchored Advantage Normalization for Multi-Source Visual Reasoning", "summary": "Hugging Face daily paper entry for Does Seeing More Mean Knowing More? Mono-Anchored Advantage Normalization for Multi-Source Visual Reasoning (2605.25437).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25437", "paper": "https://arxiv.org/abs/2605.25437", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26632", "mode": "signal", "kind": "paper", "title": "RT-Lynx: Putting the GEMM Sparsity In a Right Way for Diffusion Models", "summary": "Hugging Face daily paper entry for RT-Lynx: Putting the GEMM Sparsity In a Right Way for Diffusion Models (2605.26632).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26632", "paper": "https://arxiv.org/abs/2605.26632", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25802", "mode": "signal", "kind": "paper", "title": "Rethinking VLM Representation for VLA Initialization", "summary": "Hugging Face daily paper entry for Rethinking VLM Representation for VLA Initialization (2605.25802).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25802", "paper": "https://arxiv.org/abs/2605.25802", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27358", "mode": "signal", "kind": "paper", "title": "MobileMoE: Scaling On-Device Mixture of Experts", "summary": "Hugging Face daily paper entry for MobileMoE: Scaling On-Device Mixture of Experts (2605.27358).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.27358", "paper": "https://arxiv.org/abs/2605.27358", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26045", "mode": "signal", "kind": "paper", "title": "Confidence and Calibration of Activation Oracles for Reliable Interpretation of Language Model Internals", "summary": "Hugging Face daily paper entry for Confidence and Calibration of Activation Oracles for Reliable Interpretation of Language Model Internals (2605.26045).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26045", "paper": "https://arxiv.org/abs/2605.26045", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25188", "mode": "signal", "kind": "paper", "title": "DarkForest: Less Talk, Higher Accuracy for Multi-Agent LLMs", "summary": "Hugging Face daily paper entry for DarkForest: Less Talk, Higher Accuracy for Multi-Agent LLMs (2605.25188).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25188", "paper": "https://arxiv.org/abs/2605.25188", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26111", "mode": "signal", "kind": "paper", "title": "Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generation", "summary": "Hugging Face daily paper entry for Squeezing Capacity from Multimodal Large Language Models for Subject-driven Generation (2605.26111).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26111", "paper": "https://arxiv.org/abs/2605.26111", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24468", "mode": "signal", "kind": "paper", "title": "SAM: State-Adaptive Memory for Long-Horizon Reasoning Agent", "summary": "Hugging Face daily paper entry for SAM: State-Adaptive Memory for Long-Horizon Reasoning Agent (2605.24468).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24468", "paper": "https://arxiv.org/abs/2605.24468", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24219", "mode": "signal", "kind": "paper", "title": "Beyond Final Answers: Auditing Trajectory-Level Hallucinations in Multi-Agent Industrial Workflows", "summary": "Hugging Face daily paper entry for Beyond Final Answers: Auditing Trajectory-Level Hallucinations in Multi-Agent Industrial Workflows (2605.24219).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24219", "paper": "https://arxiv.org/abs/2605.24219", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.27235", "mode": "signal", "kind": "paper", "title": "MRT: Masked Region Transformer for Layered Image Generation and Editing at Scale", "summary": "Hugging Face daily paper entry for MRT: Masked Region Transformer for Layered Image Generation and Editing at Scale (2605.27235).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.27235", "paper": "https://arxiv.org/abs/2605.27235", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23215", "mode": "signal", "kind": "paper", "title": "FastKernels: Benchmarking GPU Kernel Generation in Production", "summary": "Hugging Face daily paper entry for FastKernels: Benchmarking GPU Kernel Generation in Production (2605.23215).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23215", "paper": "https://arxiv.org/abs/2605.23215", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18879", "mode": "signal", "kind": "paper", "title": "ZeroUnlearn: Few-Shot Knowledge Unlearning in Large Language Models", "summary": "Hugging Face daily paper entry for ZeroUnlearn: Few-Shot Knowledge Unlearning in Large Language Models (2605.18879).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.18879", "paper": "https://arxiv.org/abs/2605.18879", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22608", "mode": "signal", "kind": "paper", "title": "Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents", "summary": "Hugging Face daily paper entry for Agentic CLEAR: Automating Multi-Level Evaluation of LLM Agents (2605.22608).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22608", "paper": "https://arxiv.org/abs/2605.22608", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26293", "mode": "signal", "kind": "paper", "title": "CroCo: Cross-Lingual Contrastive Preference Tuning on Self-Generations", "summary": "Hugging Face daily paper entry for CroCo: Cross-Lingual Contrastive Preference Tuning on Self-Generations (2605.26293).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26293", "paper": "https://arxiv.org/abs/2605.26293", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24931", "mode": "signal", "kind": "paper", "title": "Learning High-Frequency Continuous Action Chunks in Latent Space", "summary": "Hugging Face daily paper entry for Learning High-Frequency Continuous Action Chunks in Latent Space (2605.24931).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24931", "paper": "https://arxiv.org/abs/2605.24931", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22769", "mode": "signal", "kind": "paper", "title": "Understanding Data Temporality Impact on Large Language Models Pre-training", "summary": "Hugging Face daily paper entry for Understanding Data Temporality Impact on Large Language Models Pre-training (2605.22769).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22769", "paper": "https://arxiv.org/abs/2605.22769", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26242", "mode": "signal", "kind": "paper", "title": "Can LLMs Introspect? A Reality Check", "summary": "Hugging Face daily paper entry for Can LLMs Introspect? A Reality Check (2605.26242).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26242", "paper": "https://arxiv.org/abs/2605.26242", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.25162", "mode": "signal", "kind": "paper", "title": "STREAM: A Data-Centric Framework for Mining High-Value Task-Oriented Dialogues from Streaming Media", "summary": "Hugging Face daily paper entry for STREAM: A Data-Centric Framework for Mining High-Value Task-Oriented Dialogues from Streaming Media (2605.25162).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25162", "paper": "https://arxiv.org/abs/2605.25162", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.15042", "mode": "signal", "kind": "paper", "title": "EverAnimate: Minute-Scale Human Animation via Latent Flow Restoration", "summary": "Hugging Face daily paper entry for EverAnimate: Minute-Scale Human Animation via Latent Flow Restoration (2605.15042).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.15042", "paper": "https://arxiv.org/abs/2605.15042", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2503.08600", "mode": "signal", "kind": "paper", "title": "NSF-SciFy: Mining the NSF Awards Database for Scientific Claims", "summary": "Hugging Face daily paper entry for NSF-SciFy: Mining the NSF Awards Database for Scientific Claims (2503.08600).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-27", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2503.08600", "paper": "https://arxiv.org/abs/2503.08600", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-27", "sourceUpdatedAt": "2026-05-27", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "hf-dataset:anisoleai/fineweb-tokenized", "mode": "catalog", "kind": "dataset", "title": "anisoleai/fineweb-tokenized", "summary": "Hugging Face dataset with 63296 downloads.", "popularity": { "value": 63296, "label": "downloads" }, "whyItMatters": "Useful to archive because it is a source-listed dataset with measurable Hugging Face usage.", "sourceId": "huggingface-datasets", "sourceName": "Hugging Face Datasets", "sourceUrl": "https://huggingface.co/api/datasets?sort=downloads&direction=-1&limit=30", "publishedAt": "2026-05-29", "category": "datasets", "subcategory": "datasets", "tags": [ "arxiv:2406.17557", "datasets", "language:en", "license:odc-by", "modality:tabular", "modality:text", "pre-training", "region:us" ], "links": { "primary": "https://huggingface.co/datasets/anisoleai/fineweb-tokenized", "paper": "", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": false, "license": "odc-by", "status": "active", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-29", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.25604", "mode": "signal", "kind": "paper", "title": "DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning", "summary": "Hugging Face daily paper entry for DVAO: Dynamic Variance-adaptive Advantage Optimization for Multi-reward Reinforcement Learning (2605.25604).", "popularity": { "value": 133, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25604", "paper": "https://arxiv.org/abs/2605.25604", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25874", "mode": "signal", "kind": "paper", "title": "WBench: A Comprehensive Multi-turn Benchmark for Interactive Video World Model Evaluation", "summary": "Hugging Face daily paper entry for WBench: A Comprehensive Multi-turn Benchmark for Interactive Video World Model Evaluation (2605.25874).", "popularity": { "value": 101, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25874", "paper": "https://arxiv.org/abs/2605.25874", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24830", "mode": "signal", "kind": "paper", "title": "Macaron-A2UI: A Model for Generative UI in Personal Agents", "summary": "Hugging Face daily paper entry for Macaron-A2UI: A Model for Generative UI in Personal Agents (2605.24830).", "popularity": { "value": 80, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24830", "paper": "https://arxiv.org/abs/2605.24830", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23218", "mode": "signal", "kind": "paper", "title": "Foundation Protocol: A Coordination Layer for Agentic Society", "summary": "Hugging Face daily paper entry for Foundation Protocol: A Coordination Layer for Agentic Society (2605.23218).", "popularity": { "value": 79, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23218", "paper": "https://arxiv.org/abs/2605.23218", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26115", "mode": "signal", "kind": "paper", "title": "TriSplat: Simulation-Ready Feed-Forward 3D Scene Reconstruction", "summary": "Hugging Face daily paper entry for TriSplat: Simulation-Ready Feed-Forward 3D Scene Reconstruction (2605.26115).", "popularity": { "value": 51, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26115", "paper": "https://arxiv.org/abs/2605.26115", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25343", "mode": "signal", "kind": "paper", "title": "Toward Native Multimodal Modeling: A Roadmap", "summary": "Hugging Face daily paper entry for Toward Native Multimodal Modeling: A Roadmap (2605.25343).", "popularity": { "value": 42, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25343", "paper": "https://arxiv.org/abs/2605.25343", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25535", "mode": "signal", "kind": "paper", "title": "Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents", "summary": "Hugging Face daily paper entry for Personalize-then-Store: Benchmarking and Learning Personalized Memory for Long-horizon Agents (2605.25535).", "popularity": { "value": 41, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25535", "paper": "https://arxiv.org/abs/2605.25535", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24218", "mode": "signal", "kind": "paper", "title": "QUEST: Training Frontier Deep Research Agents with Fully Synthetic Tasks", "summary": "Hugging Face daily paper entry for QUEST: Training Frontier Deep Research Agents with Fully Synthetic Tasks (2605.24218).", "popularity": { "value": 41, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24218", "paper": "https://arxiv.org/abs/2605.24218", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23081", "mode": "signal", "kind": "paper", "title": "ThriftAttention: Selective Mixed Precision for Long-Context FP4 Attention", "summary": "Hugging Face daily paper entry for ThriftAttention: Selective Mixed Precision for Long-Context FP4 Attention (2605.23081).", "popularity": { "value": 41, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23081", "paper": "https://arxiv.org/abs/2605.23081", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20342", "mode": "signal", "kind": "paper", "title": "ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning", "summary": "Hugging Face daily paper entry for ParaVT: Taming the Tool Prior Paradox for Parallel Tool Use in Agentic Video Reinforcement Learning (2605.20342).", "popularity": { "value": 34, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20342", "paper": "https://arxiv.org/abs/2605.20342", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25624", "mode": "signal", "kind": "paper", "title": "CUA-Gym: Scaling Verifiable Training Environments and Tasks for Computer-Use Agents", "summary": "Hugging Face daily paper entry for CUA-Gym: Scaling Verifiable Training Environments and Tasks for Computer-Use Agents (2605.25624).", "popularity": { "value": 32, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25624", "paper": "https://arxiv.org/abs/2605.25624", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23204", "mode": "signal", "kind": "paper", "title": "AutoResearch AI: Towards AI-Powered Research Automation for Scientific Discovery", "summary": "Hugging Face daily paper entry for AutoResearch AI: Towards AI-Powered Research Automation for Scientific Discovery (2605.23204).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23204", "paper": "https://arxiv.org/abs/2605.23204", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24938", "mode": "signal", "kind": "paper", "title": "Your Embedding Model is SMARTer Than You Think", "summary": "Hugging Face daily paper entry for Your Embedding Model is SMARTer Than You Think (2605.24938).", "popularity": { "value": 25, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "embeddings", "tags": [ "embeddings", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24938", "paper": "https://arxiv.org/abs/2605.24938", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26086", "mode": "signal", "kind": "paper", "title": "Claw-Anything: Benchmarking Always-On Personal Assistants with Broader Access to User's Digital World", "summary": "Hugging Face daily paper entry for Claw-Anything: Benchmarking Always-On Personal Assistants with Broader Access to User's Digital World (2605.26086).", "popularity": { "value": 23, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26086", "paper": "https://arxiv.org/abs/2605.26086", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25569", "mode": "signal", "kind": "paper", "title": "ControlLight: Towards Controllable, Consistent, and Generalizable Low-Light Enhancement", "summary": "Hugging Face daily paper entry for ControlLight: Towards Controllable, Consistent, and Generalizable Low-Light Enhancement (2605.25569).", "popularity": { "value": 21, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25569", "paper": "https://arxiv.org/abs/2605.25569", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25449", "mode": "signal", "kind": "paper", "title": "Pantheon360: Taming Digital Twin Generation via 3D-Aware 360° Video Diffusion", "summary": "Hugging Face daily paper entry for Pantheon360: Taming Digital Twin Generation via 3D-Aware 360° Video Diffusion (2605.25449).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.25449", "paper": "https://arxiv.org/abs/2605.25449", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24117", "mode": "signal", "kind": "paper", "title": "SkillEvolBench: Benchmarking the Evolution from Episodic Experience to Procedural Skills", "summary": "Hugging Face daily paper entry for SkillEvolBench: Benchmarking the Evolution from Episodic Experience to Procedural Skills (2605.24117).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24117", "paper": "https://arxiv.org/abs/2605.24117", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26535", "mode": "signal", "kind": "paper", "title": "Recursive Flow Matching", "summary": "Hugging Face daily paper entry for Recursive Flow Matching (2605.26535).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26535", "paper": "https://arxiv.org/abs/2605.26535", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26105", "mode": "signal", "kind": "paper", "title": "On-Policy Adversarial Flow Distillation for Autoregressive Video Generation", "summary": "Hugging Face daily paper entry for On-Policy Adversarial Flow Distillation for Autoregressive Video Generation (2605.26105).", "popularity": { "value": 18, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.26105", "paper": "https://arxiv.org/abs/2605.26105", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26102", "mode": "signal", "kind": "paper", "title": "InstructSAM: Segment Any Instance with Any Instructions", "summary": "Hugging Face daily paper entry for InstructSAM: Segment Any Instance with Any Instructions (2605.26102).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26102", "paper": "https://arxiv.org/abs/2605.26102", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23986", "mode": "signal", "kind": "paper", "title": "MemForest: An Efficient Agent Memory System with Hierarchical Temporal Indexing", "summary": "Hugging Face daily paper entry for MemForest: An Efficient Agent Memory System with Hierarchical Temporal Indexing (2605.23986).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23986", "paper": "https://arxiv.org/abs/2605.23986", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25971", "mode": "signal", "kind": "paper", "title": "Anticipate and Learn: Unleashing Idle-Time Compute in Proactive Agents", "summary": "Hugging Face daily paper entry for Anticipate and Learn: Unleashing Idle-Time Compute in Proactive Agents (2605.25971).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25971", "paper": "https://arxiv.org/abs/2605.25971", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26089", "mode": "signal", "kind": "paper", "title": "Channel-wise Vector Quantization", "summary": "Hugging Face daily paper entry for Channel-wise Vector Quantization (2605.26089).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26089", "paper": "https://arxiv.org/abs/2605.26089", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21748", "mode": "signal", "kind": "paper", "title": "RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Benchmark Generator", "summary": "Hugging Face daily paper entry for RankJudge: A Multi-Turn LLM-as-a-Judge Synthetic Benchmark Generator (2605.21748).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21748", "paper": "https://arxiv.org/abs/2605.21748", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26109", "mode": "signal", "kind": "paper", "title": "Helix4D: Complex 4D Mesh Generation", "summary": "Hugging Face daily paper entry for Helix4D: Complex 4D Mesh Generation (2605.26109).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26109", "paper": "https://arxiv.org/abs/2605.26109", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25052", "mode": "signal", "kind": "paper", "title": "Faithfulness Metrics Don't Measure Faithfulness: A Meta-Evaluation with Ground Truth", "summary": "Hugging Face daily paper entry for Faithfulness Metrics Don't Measure Faithfulness: A Meta-Evaluation with Ground Truth (2605.25052).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25052", "paper": "https://arxiv.org/abs/2605.25052", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24213", "mode": "signal", "kind": "paper", "title": "Towards Evaluation Engineering: An Empirical Study of ML Evaluation Harnesses in the Wild", "summary": "Hugging Face daily paper entry for Towards Evaluation Engineering: An Empirical Study of ML Evaluation Harnesses in the Wild (2605.24213).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24213", "paper": "https://arxiv.org/abs/2605.24213", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26099", "mode": "signal", "kind": "paper", "title": "Language Models Need Sleep", "summary": "Hugging Face daily paper entry for Language Models Need Sleep (2605.26099).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26099", "paper": "https://arxiv.org/abs/2605.26099", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25294", "mode": "signal", "kind": "paper", "title": "Geometry-Aware Image Flow Matching", "summary": "Hugging Face daily paper entry for Geometry-Aware Image Flow Matching (2605.25294).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.25294", "paper": "https://arxiv.org/abs/2605.25294", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24426", "mode": "signal", "kind": "paper", "title": "SEAL: Synergistic Co-Evolution of Agents and Learning Environments", "summary": "Hugging Face daily paper entry for SEAL: Synergistic Co-Evolution of Agents and Learning Environments (2605.24426).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24426", "paper": "https://arxiv.org/abs/2605.24426", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23699", "mode": "signal", "kind": "paper", "title": "CRONOS: Benchmarking Counterfactual Physical Consistency in Video Models", "summary": "Hugging Face daily paper entry for CRONOS: Benchmarking Counterfactual Physical Consistency in Video Models (2605.23699).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23699", "paper": "https://arxiv.org/abs/2605.23699", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.08129", "mode": "signal", "kind": "paper", "title": "Towards Customized Multimodal Role-Play", "summary": "Hugging Face daily paper entry for Towards Customized Multimodal Role-Play (2605.08129).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.08129", "paper": "https://arxiv.org/abs/2605.08129", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23491", "mode": "signal", "kind": "paper", "title": "CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test", "summary": "Hugging Face daily paper entry for CoSPlay: Cooperative Self-Play at Test-Time with Self-Generated Code and Unit Test (2605.23491).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23491", "paper": "https://arxiv.org/abs/2605.23491", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25461", "mode": "signal", "kind": "paper", "title": "MetaphorVU: Towards Metaphorical Video Understanding", "summary": "Hugging Face daily paper entry for MetaphorVU: Towards Metaphorical Video Understanding (2605.25461).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.25461", "paper": "https://arxiv.org/abs/2605.25461", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25160", "mode": "signal", "kind": "paper", "title": "SimuWoB: Simulating Real-World Mobile Apps for Fast and Faithful GUI Agent Benchmarking", "summary": "Hugging Face daily paper entry for SimuWoB: Simulating Real-World Mobile Apps for Fast and Faithful GUI Agent Benchmarking (2605.25160).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25160", "paper": "https://arxiv.org/abs/2605.25160", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26502", "mode": "signal", "kind": "paper", "title": "PRISM: Position-encoded Regressive Inverse Spectral Model for Multilayer Thin-Film Design", "summary": "Hugging Face daily paper entry for PRISM: Position-encoded Regressive Inverse Spectral Model for Multilayer Thin-Film Design (2605.26502).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26502", "paper": "https://arxiv.org/abs/2605.26502", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24517", "mode": "signal", "kind": "paper", "title": "ECHO: Terminal Agents Learn World Models for Free", "summary": "Hugging Face daily paper entry for ECHO: Terminal Agents Learn World Models for Free (2605.24517).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24517", "paper": "https://arxiv.org/abs/2605.24517", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23264", "mode": "signal", "kind": "paper", "title": "Coloring the Noise: Adversarial Sobolev Alignment for Faithful Image Super Resolution", "summary": "Hugging Face daily paper entry for Coloring the Noise: Adversarial Sobolev Alignment for Faithful Image Super Resolution (2605.23264).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.23264", "paper": "https://arxiv.org/abs/2605.23264", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22880", "mode": "signal", "kind": "paper", "title": "How Far Will They Go? Red-Teaming Online Influence with Large Language Models", "summary": "Hugging Face daily paper entry for How Far Will They Go? Red-Teaming Online Influence with Large Language Models (2605.22880).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22880", "paper": "https://arxiv.org/abs/2605.22880", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26108", "mode": "signal", "kind": "paper", "title": "Reinforcing Few-step Generators via Reward-Tilted Distribution Matching", "summary": "Hugging Face daily paper entry for Reinforcing Few-step Generators via Reward-Tilted Distribution Matching (2605.26108).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26108", "paper": "https://arxiv.org/abs/2605.26108", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.25191", "mode": "signal", "kind": "paper", "title": "Injecting Image Guidance into Text-Conditioned Diffusion Models at Inference", "summary": "Hugging Face daily paper entry for Injecting Image Guidance into Text-Conditioned Diffusion Models at Inference (2605.25191).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25191", "paper": "https://arxiv.org/abs/2605.25191", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2604.13517", "mode": "signal", "kind": "paper", "title": "Representation over Routing: Diagnosing Temporal Routing Pathologies in Multi-Timescale PPO", "summary": "Hugging Face daily paper entry for Representation over Routing: Diagnosing Temporal Routing Pathologies in Multi-Timescale PPO (2604.13517).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2604.13517", "paper": "https://arxiv.org/abs/2604.13517", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.26449", "mode": "signal", "kind": "paper", "title": "Cross-scale Aligned Supervision for Training GANs", "summary": "Hugging Face daily paper entry for Cross-scale Aligned Supervision for Training GANs (2605.26449).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.26449", "paper": "https://arxiv.org/abs/2605.26449", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.26095", "mode": "signal", "kind": "paper", "title": "Pixel-Level Pavement Distress Assessment Using Instance Segmentation", "summary": "Hugging Face daily paper entry for Pixel-Level Pavement Distress Assessment Using Instance Segmentation (2605.26095).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.26095", "paper": "https://arxiv.org/abs/2605.26095", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.26002", "mode": "signal", "kind": "paper", "title": "SemBridge: Language Transfer in Sparse Encoders via Multilingual Semantic Bridges", "summary": "Hugging Face daily paper entry for SemBridge: Language Transfer in Sparse Encoders via Multilingual Semantic Bridges (2605.26002).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.26002", "paper": "https://arxiv.org/abs/2605.26002", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.25189", "mode": "signal", "kind": "paper", "title": "Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models", "summary": "Hugging Face daily paper entry for Directional Alignment Mitigates Reward Hacking in Reinforcement Learning for Language Models (2605.25189).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.25189", "paper": "https://arxiv.org/abs/2605.25189", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.23889", "mode": "signal", "kind": "paper", "title": "HorizonStream: Long-Horizon Attention for Streaming 3D Reconstruction", "summary": "Hugging Face daily paper entry for HorizonStream: Long-Horizon Attention for Streaming 3D Reconstruction (2605.23889).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23889", "paper": "https://arxiv.org/abs/2605.23889", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22818", "mode": "signal", "kind": "paper", "title": "MotiMotion: Motion-Controlled Video Generation with Visual Reasoning", "summary": "Hugging Face daily paper entry for MotiMotion: Motion-Controlled Video Generation with Visual Reasoning (2605.22818).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.22818", "paper": "https://arxiv.org/abs/2605.22818", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.21712", "mode": "signal", "kind": "paper", "title": "Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries", "summary": "Hugging Face daily paper entry for Broadening Access to Transportation Safety Data with Generative AI: A Schema-Grounded Framework for Spatial Natural Language Queries (2605.21712).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21712", "paper": "https://arxiv.org/abs/2605.21712", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.21085", "mode": "signal", "kind": "paper", "title": "Decoupling Communication from Policy: Robust MARL under Bandwidth Constraints", "summary": "Hugging Face daily paper entry for Decoupling Communication from Policy: Robust MARL under Bandwidth Constraints (2605.21085).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21085", "paper": "https://arxiv.org/abs/2605.21085", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20278", "mode": "signal", "kind": "paper", "title": "ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison", "summary": "Hugging Face daily paper entry for ClaimDiff-RL: Fine-Grained Caption Reinforcement Learning through Visual Claim Comparison (2605.20278).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20278", "paper": "https://arxiv.org/abs/2605.20278", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.10988", "mode": "signal", "kind": "paper", "title": "Seeing the Needle in the Haystack: Towards Weakly-Supervised Log Instance Anomaly Localization via Counterfactual Perturbation", "summary": "Hugging Face daily paper entry for Seeing the Needle in the Haystack: Towards Weakly-Supervised Log Instance Anomaly Localization via Counterfactual Perturbation (2605.10988).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.10988", "paper": "https://arxiv.org/abs/2605.10988", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2603.16331", "mode": "signal", "kind": "paper", "title": "Decoding the Critique Mechanism in Large Reasoning Models", "summary": "Hugging Face daily paper entry for Decoding the Critique Mechanism in Large Reasoning Models (2603.16331).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-26", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2603.16331", "paper": "https://arxiv.org/abs/2603.16331", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-26", "sourceUpdatedAt": "2026-05-26", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "hf-dataset:just-me7ss/American-Sign-Language-Dataset", "mode": "catalog", "kind": "dataset", "title": "just-me7ss/American-Sign-Language-Dataset", "summary": "Hugging Face dataset with 273318 downloads.", "popularity": { "value": 273318, "label": "downloads" }, "whyItMatters": "Useful to archive because it is a source-listed dataset with measurable Hugging Face usage.", "sourceId": "huggingface-datasets", "sourceName": "Hugging Face Datasets", "sourceUrl": "https://huggingface.co/api/datasets?sort=downloads&direction=-1&limit=30", "publishedAt": "2026-05-25", "category": "datasets", "subcategory": "datasets", "tags": [ "american sign language", "asl", "datasets", "gesture recognition", "license:mit", "modality:video", "region:us", "video dataset" ], "links": { "primary": "https://huggingface.co/datasets/just-me7ss/American-Sign-Language-Dataset", "paper": "", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": false, "license": "mit", "status": "active", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.23904", "mode": "signal", "kind": "paper", "title": "SkillOpt: Executive Strategy for Self-Evolving Agent Skills", "summary": "SkillOpt introduces a systematic text-space optimizer for agent skills that trains skills as external agent state with stable updates and zero deployment inference overhead, achieving superior performance across multiple benchmarks and execution environment...", "popularity": { "value": 216, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23904", "paper": "https://arxiv.org/abs/2605.23904", "repo": "https://github.com/microsoft/SkillOpt", "docs": "https://arxiv.org/pdf/2605.23904" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20708", "mode": "signal", "kind": "paper", "title": "Rethinking Cross-Layer Information Routing in Diffusion Transformers", "summary": "Hugging Face daily paper entry for Rethinking Cross-Layer Information Routing in Diffusion Transformers (2605.20708).", "popularity": { "value": 109, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20708", "paper": "https://arxiv.org/abs/2605.20708", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21573", "mode": "signal", "kind": "paper", "title": "Lens: Rethinking Training Efficiency for Foundational Text-to-Image Models", "summary": "Hugging Face daily paper entry for Lens: Rethinking Training Efficiency for Foundational Text-to-Image Models (2605.21573).", "popularity": { "value": 107, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.21573", "paper": "https://arxiv.org/abs/2605.21573", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22878", "mode": "signal", "kind": "paper", "title": "SciAtlas: A Large-Scale Knowledge Graph for Automated Scientific Research", "summary": "Hugging Face daily paper entry for SciAtlas: A Large-Scale Knowledge Graph for Automated Scientific Research (2605.22878).", "popularity": { "value": 58, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.22878", "paper": "https://arxiv.org/abs/2605.22878", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23463", "mode": "signal", "kind": "paper", "title": "StepAudio 2.5 Technical Report", "summary": "Hugging Face daily paper entry for StepAudio 2.5 Technical Report (2605.23463).", "popularity": { "value": 49, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23463", "paper": "https://arxiv.org/abs/2605.23463", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23902", "mode": "signal", "kind": "paper", "title": "PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion", "summary": "Hugging Face daily paper entry for PiD: Fast and High-Resolution Latent Decoding with Pixel Diffusion (2605.23902).", "popularity": { "value": 45, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23902", "paper": "https://arxiv.org/abs/2605.23902", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18018", "mode": "signal", "kind": "paper", "title": "See What I Mean: Aligning Vision and Language Representations for Video Fine-grained Object Understanding", "summary": "Hugging Face daily paper entry for See What I Mean: Aligning Vision and Language Representations for Video Fine-grained Object Understanding (2605.18018).", "popularity": { "value": 33, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.18018", "paper": "https://arxiv.org/abs/2605.18018", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23899", "mode": "signal", "kind": "paper", "title": "From Raw Experience to Skill Consumption: A Systematic Study of Model-Generated Agent Skills", "summary": "Hugging Face daily paper entry for From Raw Experience to Skill Consumption: A Systematic Study of Model-Generated Agent Skills (2605.23899).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23899", "paper": "https://arxiv.org/abs/2605.23899", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23771", "mode": "signal", "kind": "paper", "title": "PhotoFlow: Agentic 3D Virtual Photography Missions", "summary": "Hugging Face daily paper entry for PhotoFlow: Agentic 3D Virtual Photography Missions (2605.23771).", "popularity": { "value": 26, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23771", "paper": "https://arxiv.org/abs/2605.23771", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22570", "mode": "signal", "kind": "paper", "title": "VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis", "summary": "Hugging Face daily paper entry for VGenST-Bench: A Benchmark for Spatio-Temporal Reasoning via Active Video Synthesis (2605.22570).", "popularity": { "value": 24, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22570", "paper": "https://arxiv.org/abs/2605.22570", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21195", "mode": "signal", "kind": "paper", "title": "RankE: End-to-End Post-Training for Discrete Text-to-Image Generation with Decoder Co-Evolution", "summary": "Hugging Face daily paper entry for RankE: End-to-End Post-Training for Discrete Text-to-Image Generation with Decoder Co-Evolution (2605.21195).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.21195", "paper": "https://arxiv.org/abs/2605.21195", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17448", "mode": "signal", "kind": "paper", "title": "Self-Improving CAD Generation Agents with Finite Element Analysis as Feedback", "summary": "Hugging Face daily paper entry for Self-Improving CAD Generation Agents with Finite Element Analysis as Feedback (2605.17448).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17448", "paper": "https://arxiv.org/abs/2605.17448", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23345", "mode": "signal", "kind": "paper", "title": "SCOPE: Simulating Cross-game Operations in Playable Environments for FPS World Models", "summary": "Hugging Face daily paper entry for SCOPE: Simulating Cross-game Operations in Playable Environments for FPS World Models (2605.23345).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23345", "paper": "https://arxiv.org/abs/2605.23345", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23901", "mode": "signal", "kind": "paper", "title": "LLMs as Noisy Channels: A Shannon Perspective on Model Capacity and Scaling Laws", "summary": "Hugging Face daily paper entry for LLMs as Noisy Channels: A Shannon Perspective on Model Capacity and Scaling Laws (2605.23901).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23901", "paper": "https://arxiv.org/abs/2605.23901", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23897", "mode": "signal", "kind": "paper", "title": "ETCHR: Editing To Clarify and Harness Reasoning", "summary": "Hugging Face daily paper entry for ETCHR: Editing To Clarify and Harness Reasoning (2605.23897).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23897", "paper": "https://arxiv.org/abs/2605.23897", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23888", "mode": "signal", "kind": "paper", "title": "GenRecon: Bridging Generative Priors for Multi-View 3D Scene Reconstruction", "summary": "Hugging Face daily paper entry for GenRecon: Bridging Generative Priors for Multi-View 3D Scene Reconstruction (2605.23888).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23888", "paper": "https://arxiv.org/abs/2605.23888", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17873", "mode": "signal", "kind": "paper", "title": "HINT-SD: Targeted Hindsight Self-Distillation for Long-Horizon Agents", "summary": "Hugging Face daily paper entry for HINT-SD: Targeted Hindsight Self-Distillation for Long-Horizon Agents (2605.17873).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17873", "paper": "https://arxiv.org/abs/2605.17873", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23903", "mode": "signal", "kind": "paper", "title": "Geo-Align: Video Generation Alignment via Metric Geometry Reward", "summary": "Hugging Face daily paper entry for Geo-Align: Video Generation Alignment via Metric Geometry Reward (2605.23903).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.23903", "paper": "https://arxiv.org/abs/2605.23903", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20177", "mode": "signal", "kind": "paper", "title": "From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models", "summary": "Hugging Face daily paper entry for From Seeing to Thinking: Decoupling Perception and Reasoning Improves Post-Training of Vision-Language Models (2605.20177).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20177", "paper": "https://arxiv.org/abs/2605.20177", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.23892", "mode": "signal", "kind": "paper", "title": "Good Token Hunting: A Hitchhiker's Guide to Token Selection for Visual Geometry Transformers", "summary": "Hugging Face daily paper entry for Good Token Hunting: A Hitchhiker's Guide to Token Selection for Visual Geometry Transformers (2605.23892).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.23892", "paper": "https://arxiv.org/abs/2605.23892", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21856", "mode": "signal", "kind": "paper", "title": "The Illusion of Reasoning: Exposing Evasive Data Contamination in LLMs via Zero-CoT Truncation", "summary": "Hugging Face daily paper entry for The Illusion of Reasoning: Exposing Evasive Data Contamination in LLMs via Zero-CoT Truncation (2605.21856).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21856", "paper": "https://arxiv.org/abs/2605.21856", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19282", "mode": "signal", "kind": "paper", "title": "Rethinking Muon Beyond Pretraining: Spectral Failures and High-Pass Remedies for VLA and RLVR", "summary": "Hugging Face daily paper entry for Rethinking Muon Beyond Pretraining: Spectral Failures and High-Pass Remedies for VLA and RLVR (2605.19282).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19282", "paper": "https://arxiv.org/abs/2605.19282", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17766", "mode": "signal", "kind": "paper", "title": "LatentUMM: Dual Latent Alignment for Unified Multimodal Models", "summary": "Hugging Face daily paper entry for LatentUMM: Dual Latent Alignment for Unified Multimodal Models (2605.17766).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17766", "paper": "https://arxiv.org/abs/2605.17766", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2604.20665", "mode": "signal", "kind": "paper", "title": "The Expense of Seeing: Attaining Trustworthy Multimodal Reasoning Within the Monolithic Paradigm", "summary": "Hugging Face daily paper entry for The Expense of Seeing: Attaining Trustworthy Multimodal Reasoning Within the Monolithic Paradigm (2604.20665).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2604.20665", "paper": "https://arxiv.org/abs/2604.20665", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21488", "mode": "signal", "kind": "paper", "title": "Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning", "summary": "Hugging Face daily paper entry for Equilibrium Reasoners: Learning Attractors Enables Scalable Reasoning (2605.21488).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21488", "paper": "https://arxiv.org/abs/2605.21488", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24681", "mode": "signal", "kind": "paper", "title": "Mix-MoE: Improving Multilingual Machine Translation of Large Language Models through Mixed MoEs", "summary": "Hugging Face daily paper entry for Mix-MoE: Improving Multilingual Machine Translation of Large Language Models through Mixed MoEs (2605.24681).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.24681", "paper": "https://arxiv.org/abs/2605.24681", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.24675", "mode": "signal", "kind": "paper", "title": "VaaWIT: Visual-Aware Adaptation of Large Language Models for Multilingual Web Image Translation", "summary": "Hugging Face daily paper entry for VaaWIT: Visual-Aware Adaptation of Large Language Models for Multilingual Web Image Translation (2605.24675).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.24675", "paper": "https://arxiv.org/abs/2605.24675", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.19354", "mode": "signal", "kind": "paper", "title": "Next-Acceleration-Scale Prediction for Autoregressive MRI Reconstruction", "summary": "Hugging Face daily paper entry for Next-Acceleration-Scale Prediction for Autoregressive MRI Reconstruction (2605.19354).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-25", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19354", "paper": "https://arxiv.org/abs/2605.19354", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-25", "sourceUpdatedAt": "2026-05-25", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.21467", "mode": "signal", "kind": "paper", "title": "DelTA: Discriminative Token Credit Assignment for Reinforcement Learning from Verifiable Rewards", "summary": "Reinforcement learning from verifiable rewards is enhanced through a discriminative token credit assignment method that improves reward-based training by amplifying distinctive token-gradient directions and reducing noise from shared patterns. Reinforcement...", "popularity": { "value": 204, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21467", "paper": "https://arxiv.org/abs/2605.21467", "repo": "https://github.com/RUCBM/DelTA", "docs": "https://arxiv.org/pdf/2605.21467" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22355", "mode": "signal", "kind": "paper", "title": "TransitLM: A Large-Scale Dataset and Benchmark for Map-Free Transit Route Generation", "summary": "TransitLM dataset enables end-to-end transit route planning using large language models trained on structured transit data, eliminating the need for traditional map-based approaches. Public transit route planning traditionally depends on structured map infr...", "popularity": { "value": 174, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22355", "paper": "https://arxiv.org/abs/2605.22355", "repo": "https://github.com/HotTricker/TransitLM", "docs": "https://arxiv.org/pdf/2605.22355" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22109", "mode": "signal", "kind": "paper", "title": "Perception or Prejudice: Can MLLMs Go Beyond First Impressions of Personality?", "summary": "Researchers introduce a new task and dataset for evaluating personality reasoning in multimodal language models, revealing significant gaps between accurate predictions and grounded reasoning processes. Multimodal Large Language Models (MLLMs) are increasin...", "popularity": { "value": 169, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22109", "paper": "https://arxiv.org/abs/2605.22109", "repo": "https://github.com/kkkcx/MM-OCEAN", "docs": "https://arxiv.org/pdf/2605.22109" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14678", "mode": "signal", "kind": "paper", "title": "π-Bench: Evaluating Proactive Personal Assistant Agents in Long-Horizon Workflows", "summary": "Hugging Face daily paper entry for π-Bench: Evaluating Proactive Personal Assistant Agents in Long-Horizon Workflows (2605.14678).", "popularity": { "value": 102, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.14678", "paper": "https://arxiv.org/abs/2605.14678", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16928", "mode": "signal", "kind": "paper", "title": "Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps", "summary": "Hugging Face daily paper entry for Full Attention Strikes Back: Transferring Full Attention into Sparse within Hundred Training Steps (2605.16928).", "popularity": { "value": 93, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16928", "paper": "https://arxiv.org/abs/2605.16928", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21850", "mode": "signal", "kind": "paper", "title": "ACC: Compiling Agent Trajectories for Long-Context Training", "summary": "Hugging Face daily paper entry for ACC: Compiling Agent Trajectories for Long-Context Training (2605.21850).", "popularity": { "value": 58, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21850", "paper": "https://arxiv.org/abs/2605.21850", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21572", "mode": "signal", "kind": "paper", "title": "PhysX-Omni: Unified Simulation-Ready Physical 3D Generation for Rigid, Deformable, and Articulated Objects", "summary": "Hugging Face daily paper entry for PhysX-Omni: Unified Simulation-Ready Physical 3D Generation for Rigid, Deformable, and Articulated Objects (2605.21572).", "popularity": { "value": 52, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21572", "paper": "https://arxiv.org/abs/2605.21572", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22012", "mode": "signal", "kind": "paper", "title": "LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning", "summary": "Hugging Face daily paper entry for LatentOmni: Rethinking Omni-Modal Understanding via Unified Audio-Visual Latent Reasoning (2605.22012).", "popularity": { "value": 46, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22012", "paper": "https://arxiv.org/abs/2605.22012", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22681", "mode": "signal", "kind": "paper", "title": "Forecasting Scientific Progress with Artificial Intelligence", "summary": "Hugging Face daily paper entry for Forecasting Scientific Progress with Artificial Intelligence (2605.22681).", "popularity": { "value": 43, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22681", "paper": "https://arxiv.org/abs/2605.22681", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22718", "mode": "signal", "kind": "paper", "title": "WorldKV: Efficient World Memory with World Retrieval and Compression", "summary": "Hugging Face daily paper entry for WorldKV: Efficient World Memory with World Retrieval and Compression (2605.22718).", "popularity": { "value": 41, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22718", "paper": "https://arxiv.org/abs/2605.22718", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22668", "mode": "signal", "kind": "paper", "title": "SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformers", "summary": "Hugging Face daily paper entry for SEGA: Spectral-Energy Guided Attention for Resolution Extrapolation in Diffusion Transformers (2605.22668).", "popularity": { "value": 40, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22668", "paper": "https://arxiv.org/abs/2605.22668", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22642", "mode": "signal", "kind": "paper", "title": "Spreadsheet-RL: Advancing Large Language Model Agents on Realistic Spreadsheet Tasks via Reinforcement Learning", "summary": "Hugging Face daily paper entry for Spreadsheet-RL: Advancing Large Language Model Agents on Realistic Spreadsheet Tasks via Reinforcement Learning (2605.22642).", "popularity": { "value": 35, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22642", "paper": "https://arxiv.org/abs/2605.22642", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22791", "mode": "signal", "kind": "paper", "title": "Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention", "summary": "Hugging Face daily paper entry for Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention (2605.22791).", "popularity": { "value": 30, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22791", "paper": "https://arxiv.org/abs/2605.22791", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20910", "mode": "signal", "kind": "paper", "title": "FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching", "summary": "Hugging Face daily paper entry for FlowLong: Inference-time Long Video Generation via Manifold-constrained Tweedie Matching (2605.20910).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20910", "paper": "https://arxiv.org/abs/2605.20910", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22536", "mode": "signal", "kind": "paper", "title": "SpaceDG: Benchmarking Spatial Intelligence under Visual Degradation", "summary": "Hugging Face daily paper entry for SpaceDG: Benchmarking Spatial Intelligence under Visual Degradation (2605.22536).", "popularity": { "value": 28, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22536", "paper": "https://arxiv.org/abs/2605.22536", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22809", "mode": "signal", "kind": "paper", "title": "Sensor2Sensor: Cross-Embodiment Sensor Conversion for Autonomous Driving", "summary": "Hugging Face daily paper entry for Sensor2Sensor: Cross-Embodiment Sensor Conversion for Autonomous Driving (2605.22809).", "popularity": { "value": 27, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22809", "paper": "https://arxiv.org/abs/2605.22809", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.10158", "mode": "signal", "kind": "paper", "title": "Unsupervised Process Reward Models", "summary": "Hugging Face daily paper entry for Unsupervised Process Reward Models (2605.10158).", "popularity": { "value": 26, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.10158", "paper": "https://arxiv.org/abs/2605.10158", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21072", "mode": "signal", "kind": "paper", "title": "Q-ARVD: Quantizing Autoregressive Video Diffusion Models", "summary": "Hugging Face daily paper entry for Q-ARVD: Quantizing Autoregressive Video Diffusion Models (2605.21072).", "popularity": { "value": 21, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.21072", "paper": "https://arxiv.org/abs/2605.21072", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22177", "mode": "signal", "kind": "paper", "title": "Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles", "summary": "Hugging Face daily paper entry for Maestro: Reinforcement Learning to Orchestrate Hierarchical Model-Skill Ensembles (2605.22177).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22177", "paper": "https://arxiv.org/abs/2605.22177", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17602", "mode": "signal", "kind": "paper", "title": "AutoRubric-T2I: Robust Rule-Based Reward Model for Text-to-Image Alignment", "summary": "Hugging Face daily paper entry for AutoRubric-T2I: Robust Rule-Based Reward Model for Text-to-Image Alignment (2605.17602).", "popularity": { "value": 19, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.17602", "paper": "https://arxiv.org/abs/2605.17602", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.12817", "mode": "signal", "kind": "paper", "title": "Training Large Language Models to Predict Clinical Events", "summary": "Hugging Face daily paper entry for Training Large Language Models to Predict Clinical Events (2605.12817).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.12817", "paper": "https://arxiv.org/abs/2605.12817", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22344", "mode": "signal", "kind": "paper", "title": "Bernini: Latent Semantic Planning for Video Diffusion", "summary": "Hugging Face daily paper entry for Bernini: Latent Semantic Planning for Video Diffusion (2605.22344).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22344", "paper": "https://arxiv.org/abs/2605.22344", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18607", "mode": "signal", "kind": "paper", "title": "Forecasting Downstream Performance of LLMs With Proxy Metrics", "summary": "Hugging Face daily paper entry for Forecasting Downstream Performance of LLMs With Proxy Metrics (2605.18607).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18607", "paper": "https://arxiv.org/abs/2605.18607", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21605", "mode": "signal", "kind": "paper", "title": "GenEvolve: Self-Evolving Image Generation Agents via Tool-Orchestrated Visual Experience Distillation", "summary": "Hugging Face daily paper entry for GenEvolve: Self-Evolving Image Generation Agents via Tool-Orchestrated Visual Experience Distillation (2605.21605).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21605", "paper": "https://arxiv.org/abs/2605.21605", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20176", "mode": "signal", "kind": "paper", "title": "ClinSeekAgent: Automating Multimodal Evidence Seeking for Agentic Clinical Reasoning", "summary": "Hugging Face daily paper entry for ClinSeekAgent: Automating Multimodal Evidence Seeking for Agentic Clinical Reasoning (2605.20176).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20176", "paper": "https://arxiv.org/abs/2605.20176", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.13734", "mode": "signal", "kind": "paper", "title": "KVServe: Service-Aware KV Cache Compression for Communication-Efficient Disaggregated LLM Serving", "summary": "Hugging Face daily paper entry for KVServe: Service-Aware KV Cache Compression for Communication-Efficient Disaggregated LLM Serving (2605.13734).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.13734", "paper": "https://arxiv.org/abs/2605.13734", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22678", "mode": "signal", "kind": "paper", "title": "Swift Sampling: Selecting Temporal Surprises via Taylor Series", "summary": "Hugging Face daily paper entry for Swift Sampling: Selecting Temporal Surprises via Taylor Series (2605.22678).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22678", "paper": "https://arxiv.org/abs/2605.22678", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22138", "mode": "signal", "kind": "paper", "title": "Efficient Agentic Reasoning Through Self-Regulated Simulative Planning", "summary": "Hugging Face daily paper entry for Efficient Agentic Reasoning Through Self-Regulated Simulative Planning (2605.22138).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22138", "paper": "https://arxiv.org/abs/2605.22138", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22144", "mode": "signal", "kind": "paper", "title": "One Sentence, One Drama: Personalized Short-Form Drama Generation via Multi-Agent Systems", "summary": "Hugging Face daily paper entry for One Sentence, One Drama: Personalized Short-Form Drama Generation via Multi-Agent Systems (2605.22144).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22144", "paper": "https://arxiv.org/abs/2605.22144", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22535", "mode": "signal", "kind": "paper", "title": "TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks", "summary": "Hugging Face daily paper entry for TerminalWorld: Benchmarking Agents on Real-World Terminal Tasks (2605.22535).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22535", "paper": "https://arxiv.org/abs/2605.22535", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.13163", "mode": "signal", "kind": "paper", "title": "LoREnc: Low-Rank Encryption for Securing Foundation Models and LoRA Adapters", "summary": "Hugging Face daily paper entry for LoREnc: Low-Rank Encryption for Securing Foundation Models and LoRA Adapters (2605.13163).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.13163", "paper": "https://arxiv.org/abs/2605.13163", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22581", "mode": "signal", "kind": "paper", "title": "SceneAligner: 3D-Grounded Floorplan Localization in the Wild", "summary": "Hugging Face daily paper entry for SceneAligner: 3D-Grounded Floorplan Localization in the Wild (2605.22581).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.22581", "paper": "https://arxiv.org/abs/2605.22581", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15669", "mode": "signal", "kind": "paper", "title": "Rule2DRC: Benchmarking LLM Agents for DRC Script Synthesis with Execution-Guided Test Generation", "summary": "Hugging Face daily paper entry for Rule2DRC: Benchmarking LLM Agents for DRC Script Synthesis with Execution-Guided Test Generation (2605.15669).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15669", "paper": "https://arxiv.org/abs/2605.15669", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22766", "mode": "signal", "kind": "paper", "title": "Diversed Model Discovery via Structured Table Discovery", "summary": "Hugging Face daily paper entry for Diversed Model Discovery via Structured Table Discovery (2605.22766).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22766", "paper": "https://arxiv.org/abs/2605.22766", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22538", "mode": "signal", "kind": "paper", "title": "Segment Anything with Motion, Geometry, and Semantic Adaptation for Complex Nonlinear Visual Object Tracking", "summary": "Hugging Face daily paper entry for Segment Anything with Motion, Geometry, and Semantic Adaptation for Complex Nonlinear Visual Object Tracking (2605.22538).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22538", "paper": "https://arxiv.org/abs/2605.22538", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21363", "mode": "signal", "kind": "paper", "title": "\"I didn't Make the Micro Decisions\": Measuring, Inducing, and Exposing Goal-Level AI Contributions in Collaboration", "summary": "Hugging Face daily paper entry for \"I didn't Make the Micro Decisions\": Measuring, Inducing, and Exposing Goal-Level AI Contributions in Collaboration (2605.21363).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21363", "paper": "https://arxiv.org/abs/2605.21363", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22777", "mode": "signal", "kind": "paper", "title": "DecQ: Detail-Condensing Queries for Enhanced Reconstruction and Generation in Representation Autoencoders", "summary": "Hugging Face daily paper entry for DecQ: Detail-Condensing Queries for Enhanced Reconstruction and Generation in Representation Autoencoders (2605.22777).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22777", "paper": "https://arxiv.org/abs/2605.22777", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18577", "mode": "signal", "kind": "paper", "title": "OmniPro: A Comprehensive Benchmark for Omni-Proactive Streaming Video Understanding", "summary": "Hugging Face daily paper entry for OmniPro: A Comprehensive Benchmark for Omni-Proactive Streaming Video Understanding (2605.18577).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18577", "paper": "https://arxiv.org/abs/2605.18577", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.22717", "mode": "signal", "kind": "paper", "title": "Live Music Diffusion Models: Efficient Fine-Tuning and Post-Training of Interactive Diffusion Music Generators", "summary": "Hugging Face daily paper entry for Live Music Diffusion Models: Efficient Fine-Tuning and Post-Training of Interactive Diffusion Music Generators (2605.22717).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22717", "paper": "https://arxiv.org/abs/2605.22717", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22715", "mode": "signal", "kind": "paper", "title": "AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild", "summary": "Hugging Face daily paper entry for AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild (2605.22715).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22715", "paper": "https://arxiv.org/abs/2605.22715", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22641", "mode": "signal", "kind": "paper", "title": "More Context, Larger Models, or Moral Knowledge? A Systematic Study of Schwartz Value Detection in Political Texts", "summary": "Hugging Face daily paper entry for More Context, Larger Models, or Moral Knowledge? A Systematic Study of Schwartz Value Detection in Political Texts (2605.22641).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.22641", "paper": "https://arxiv.org/abs/2605.22641", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22552", "mode": "signal", "kind": "paper", "title": "FashionLens: Toward Versatile Fashion Image Retrieval via Task-Adaptive Learning", "summary": "Hugging Face daily paper entry for FashionLens: Toward Versatile Fashion Image Retrieval via Task-Adaptive Learning (2605.22552).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22552", "paper": "https://arxiv.org/abs/2605.22552", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.22074", "mode": "signal", "kind": "paper", "title": "From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning", "summary": "Hugging Face daily paper entry for From Reasoning Chains to Verifiable Subproblems: Curriculum Reinforcement Learning Enables Credit Assignment for LLM Reasoning (2605.22074).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.22074", "paper": "https://arxiv.org/abs/2605.22074", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.21803", "mode": "signal", "kind": "paper", "title": "Same Architecture, Different Capacity: Optimizer-Induced Spectral Scaling Laws", "summary": "Hugging Face daily paper entry for Same Architecture, Different Capacity: Optimizer-Induced Spectral Scaling Laws (2605.21803).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21803", "paper": "https://arxiv.org/abs/2605.21803", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20496", "mode": "signal", "kind": "paper", "title": "Platonic Representations in the Human Brain: Unsupervised Recovery of Universal Geometry", "summary": "Hugging Face daily paper entry for Platonic Representations in the Human Brain: Unsupervised Recovery of Universal Geometry (2605.20496).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20496", "paper": "https://arxiv.org/abs/2605.20496", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20405", "mode": "signal", "kind": "paper", "title": "Disentangling Sampling from Training Budget in Class-Imbalanced CT Body Composition Segmentation", "summary": "Hugging Face daily paper entry for Disentangling Sampling from Training Budget in Class-Imbalanced CT Body Composition Segmentation (2605.20405).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.20405", "paper": "https://arxiv.org/abs/2605.20405", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20244", "mode": "signal", "kind": "paper", "title": "Lean Refactor: Multi-Objective Controllable Proof Optimization via Agentic Strategy Search", "summary": "Hugging Face daily paper entry for Lean Refactor: Multi-Objective Controllable Proof Optimization via Agentic Strategy Search (2605.20244).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20244", "paper": "https://arxiv.org/abs/2605.20244", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.19990", "mode": "signal", "kind": "paper", "title": "Minimalist Visual Inertial Odometry", "summary": "Hugging Face daily paper entry for Minimalist Visual Inertial Odometry (2605.19990).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19990", "paper": "https://arxiv.org/abs/2605.19990", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.07604", "mode": "signal", "kind": "paper", "title": "SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild", "summary": "Hugging Face daily paper entry for SAM 3D Animal: Promptable Animal 3D Reconstruction from Images in the Wild (2605.07604).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-24", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.07604", "paper": "https://arxiv.org/abs/2605.07604", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-24", "sourceUpdatedAt": "2026-05-24", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "hf-dataset:idacy/control-arena-persistent-state-eval-logs-2026-05-23", "mode": "catalog", "kind": "dataset", "title": "idacy/control-arena-persistent-state-eval-logs-2026-05-23", "summary": "Hugging Face dataset with 47554 downloads.", "popularity": { "value": 47554, "label": "downloads" }, "whyItMatters": "Useful to archive because it is a source-listed dataset with measurable Hugging Face usage.", "sourceId": "huggingface-datasets", "sourceName": "Hugging Face Datasets", "sourceUrl": "https://huggingface.co/api/datasets?sort=downloads&direction=-1&limit=30", "publishedAt": "2026-05-23", "category": "datasets", "subcategory": "datasets", "tags": [ "datasets", "region:us" ], "links": { "primary": "https://huggingface.co/datasets/idacy/control-arena-persistent-state-eval-logs-2026-05-23", "paper": "", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": false, "license": "unknown", "status": "active", "createdAt": "2026-05-23", "sourceUpdatedAt": "2026-05-23", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14747", "mode": "signal", "kind": "paper", "title": "Video2GUI: Synthesizing Large-Scale Interaction Trajectories for Generalized GUI Agent Pretraining", "summary": "Hugging Face daily paper entry for Video2GUI: Synthesizing Large-Scale Interaction Trajectories for Generalized GUI Agent Pretraining (2605.14747).", "popularity": { "value": 145, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.14747", "paper": "https://arxiv.org/abs/2605.14747", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20613", "mode": "signal", "kind": "paper", "title": "HRM-Text: Efficient Pretraining Beyond Scaling", "summary": "Hugging Face daily paper entry for HRM-Text: Efficient Pretraining Beyond Scaling (2605.20613).", "popularity": { "value": 131, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.20613", "paper": "https://arxiv.org/abs/2605.20613", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19833", "mode": "signal", "kind": "paper", "title": "Mega-ASR: Towards In-the-wild^2 Speech Recognition via Scaling up Real-world Acoustic Simulation", "summary": "Hugging Face daily paper entry for Mega-ASR: Towards In-the-wild^2 Speech Recognition via Scaling up Real-world Acoustic Simulation (2605.19833).", "popularity": { "value": 131, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19833", "paper": "https://arxiv.org/abs/2605.19833", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18233", "mode": "signal", "kind": "paper", "title": "Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos", "summary": "Hugging Face daily paper entry for Enhancing Train-Free Infinite-Frame Generation for Consistent Long Videos (2605.18233).", "popularity": { "value": 92, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.18233", "paper": "https://arxiv.org/abs/2605.18233", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20682", "mode": "signal", "kind": "paper", "title": "IndusAgent: Reinforcing Open-Vocabulary Industrial Anomaly Detection with Agentic Tools", "summary": "Hugging Face daily paper entry for IndusAgent: Reinforcing Open-Vocabulary Industrial Anomaly Detection with Agentic Tools (2605.20682).", "popularity": { "value": 83, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20682", "paper": "https://arxiv.org/abs/2605.20682", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20266", "mode": "signal", "kind": "paper", "title": "A Survey of Large Audio Language Models: Generalization, Trustworthiness, and Outlook", "summary": "Hugging Face daily paper entry for A Survey of Large Audio Language Models: Generalization, Trustworthiness, and Outlook (2605.20266).", "popularity": { "value": 56, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20266", "paper": "https://arxiv.org/abs/2605.20266", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21468", "mode": "signal", "kind": "paper", "title": "You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectories", "summary": "Hugging Face daily paper entry for You Only Need Minimal RLVR Training: Extrapolating LLMs via Rank-1 Trajectories (2605.21468).", "popularity": { "value": 49, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21468", "paper": "https://arxiv.org/abs/2605.21468", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20873", "mode": "signal", "kind": "paper", "title": "PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models", "summary": "Hugging Face daily paper entry for PlanningBench: Generating Scalable and Verifiable Planning Data for Evaluating and Training Large Language Models (2605.20873).", "popularity": { "value": 43, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20873", "paper": "https://arxiv.org/abs/2605.20873", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19660", "mode": "signal", "kind": "paper", "title": "OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond", "summary": "Hugging Face daily paper entry for OScaR: The Occam's Razor for Extreme KV Cache Quantization in LLMs and Beyond (2605.19660).", "popularity": { "value": 40, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19660", "paper": "https://arxiv.org/abs/2605.19660", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20119", "mode": "signal", "kind": "paper", "title": "Toto 2.0: Time Series Forecasting Enters the Scaling Era", "summary": "Hugging Face daily paper entry for Toto 2.0: Time Series Forecasting Enters the Scaling Era (2605.20119).", "popularity": { "value": 38, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20119", "paper": "https://arxiv.org/abs/2605.20119", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20258", "mode": "signal", "kind": "paper", "title": "It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs", "summary": "Hugging Face daily paper entry for It Takes Two: Complementary Self-Distillation for Contextual Integrity in LLMs (2605.20258).", "popularity": { "value": 30, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20258", "paper": "https://arxiv.org/abs/2605.20258", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19376", "mode": "signal", "kind": "paper", "title": "Generative Recursive Reasoning", "summary": "Hugging Face daily paper entry for Generative Recursive Reasoning (2605.19376).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19376", "paper": "https://arxiv.org/abs/2605.19376", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20315", "mode": "signal", "kind": "paper", "title": "Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs", "summary": "Hugging Face daily paper entry for Mix-Quant: Quantized Prefilling, Precise Decoding for Agentic LLMs (2605.20315).", "popularity": { "value": 28, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20315", "paper": "https://arxiv.org/abs/2605.20315", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21487", "mode": "signal", "kind": "paper", "title": "Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning", "summary": "Hugging Face daily paper entry for Uni-Edit: Intelligent Editing Is A General Task For Unified Model Tuning (2605.21487).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21487", "paper": "https://arxiv.org/abs/2605.21487", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19484", "mode": "signal", "kind": "paper", "title": "CutVerse: A Compositional GUI Agents Benchmark for Media Post-Production Editing", "summary": "Hugging Face daily paper entry for CutVerse: A Compositional GUI Agents Benchmark for Media Post-Production Editing (2605.19484).", "popularity": { "value": 21, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19484", "paper": "https://arxiv.org/abs/2605.19484", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19597", "mode": "signal", "kind": "paper", "title": "LLMEval-Logic: A Solver-Verified Chinese Benchmark for Logical Reasoning of LLMs with Adversarial Hardening", "summary": "Hugging Face daily paper entry for LLMEval-Logic: A Solver-Verified Chinese Benchmark for Logical Reasoning of LLMs with Adversarial Hardening (2605.19597).", "popularity": { "value": 20, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19597", "paper": "https://arxiv.org/abs/2605.19597", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17991", "mode": "signal", "kind": "paper", "title": "Stable Audio 3", "summary": "Hugging Face daily paper entry for Stable Audio 3 (2605.17991).", "popularity": { "value": 18, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17991", "paper": "https://arxiv.org/abs/2605.17991", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20668", "mode": "signal", "kind": "paper", "title": "On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists", "summary": "Hugging Face daily paper entry for On the limits and opportunities of AI reviewers: Reviewing the reviews of Nature-family papers with 45 expert scientists (2605.20668).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20668", "paper": "https://arxiv.org/abs/2605.20668", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20630", "mode": "signal", "kind": "paper", "title": "Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines", "summary": "Hugging Face daily paper entry for Evaluating Temporal Semantic Caching and Workflow Optimization in Agentic Plan-Execute Pipelines (2605.20630).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20630", "paper": "https://arxiv.org/abs/2605.20630", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19804", "mode": "signal", "kind": "paper", "title": "Stitched Value Model for Diffusion Alignment", "summary": "Hugging Face daily paper entry for Stitched Value Model for Diffusion Alignment (2605.19804).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19804", "paper": "https://arxiv.org/abs/2605.19804", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2604.27263", "mode": "signal", "kind": "paper", "title": "Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation", "summary": "Hugging Face daily paper entry for Decoupling the Benefits of Subword Tokenization for Language Model Training via Byte-level Simulation (2604.27263).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2604.27263", "paper": "https://arxiv.org/abs/2604.27263", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21981", "mode": "signal", "kind": "paper", "title": "RiT: Vanilla Diffusion Transformers Suffice in Representation Space", "summary": "Hugging Face daily paper entry for RiT: Vanilla Diffusion Transformers Suffice in Representation Space (2605.21981).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21981", "paper": "https://arxiv.org/abs/2605.21981", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15113", "mode": "signal", "kind": "paper", "title": "Learning from Language Feedback via Variational Policy Distillation", "summary": "Hugging Face daily paper entry for Learning from Language Feedback via Variational Policy Distillation (2605.15113).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15113", "paper": "https://arxiv.org/abs/2605.15113", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21226", "mode": "signal", "kind": "paper", "title": "OCTOPUS: Optimized KV Cache for Transformers via Octahedral Parametrization Under optimal Squared error quantization", "summary": "Hugging Face daily paper entry for OCTOPUS: Optimized KV Cache for Transformers via Octahedral Parametrization Under optimal Squared error quantization (2605.21226).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21226", "paper": "https://arxiv.org/abs/2605.21226", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21343", "mode": "signal", "kind": "paper", "title": "OcclusionFormer: Arranging Z-Order for Layout-Grounded Image Generation", "summary": "Hugging Face daily paper entry for OcclusionFormer: Arranging Z-Order for Layout-Grounded Image Generation (2605.21343).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.21343", "paper": "https://arxiv.org/abs/2605.21343", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21131", "mode": "signal", "kind": "paper", "title": "UniT: Unified Geometry Learning with Group Autoregressive Transformer", "summary": "Hugging Face daily paper entry for UniT: Unified Geometry Learning with Group Autoregressive Transformer (2605.21131).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21131", "paper": "https://arxiv.org/abs/2605.21131", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19330", "mode": "signal", "kind": "paper", "title": "MOCHA: Multi-Objective Chebyshev Annealing for Agent Skill Optimization", "summary": "Hugging Face daily paper entry for MOCHA: Multi-Objective Chebyshev Annealing for Agent Skill Optimization (2605.19330).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19330", "paper": "https://arxiv.org/abs/2605.19330", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17526", "mode": "signal", "kind": "paper", "title": "SaaSBench: Exploring the Boundaries of Coding Agents in Long-Horizon Enterprise SaaS Engineering", "summary": "Hugging Face daily paper entry for SaaSBench: Exploring the Boundaries of Coding Agents in Long-Horizon Enterprise SaaS Engineering (2605.17526).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17526", "paper": "https://arxiv.org/abs/2605.17526", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21463", "mode": "signal", "kind": "paper", "title": "Mem-π: Adaptive Memory through Learning When and What to Generate", "summary": "Hugging Face daily paper entry for Mem-π: Adaptive Memory through Learning When and What to Generate (2605.21463).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21463", "paper": "https://arxiv.org/abs/2605.21463", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16787", "mode": "signal", "kind": "paper", "title": "The Unlearnability Phenomenon in RLVR for Language Models", "summary": "Hugging Face daily paper entry for The Unlearnability Phenomenon in RLVR for Language Models (2605.16787).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16787", "paper": "https://arxiv.org/abs/2605.16787", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20834", "mode": "signal", "kind": "paper", "title": "Conditional Equivalence of DPO and RLHF: Implicit Assumption, Failure Modes, and Provable Alignment", "summary": "Hugging Face daily paper entry for Conditional Equivalence of DPO and RLHF: Implicit Assumption, Failure Modes, and Provable Alignment (2605.20834).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20834", "paper": "https://arxiv.org/abs/2605.20834", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18565", "mode": "signal", "kind": "paper", "title": "MINTEval: Evaluating Memory under Multi-Target Interference in Long-Horizon Agent Systems", "summary": "Hugging Face daily paper entry for MINTEval: Evaluating Memory under Multi-Target Interference in Long-Horizon Agent Systems (2605.18565).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18565", "paper": "https://arxiv.org/abs/2605.18565", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.21431", "mode": "signal", "kind": "paper", "title": "iTryOn: Mastering Interactive Video Virtual Try-On with Spatial-Semantic Guidance", "summary": "Hugging Face daily paper entry for iTryOn: Mastering Interactive Video Virtual Try-On with Spatial-Semantic Guidance (2605.21431).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.21431", "paper": "https://arxiv.org/abs/2605.21431", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.21384", "mode": "signal", "kind": "paper", "title": "SpecBench: Measuring Reward Hacking in Long-Horizon Coding Agents", "summary": "Hugging Face daily paper entry for SpecBench: Measuring Reward Hacking in Long-Horizon Coding Agents (2605.21384).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.21384", "paper": "https://arxiv.org/abs/2605.21384", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20955", "mode": "signal", "kind": "paper", "title": "DrawMotion: Generating 3D Human Motions by Freehand Drawing", "summary": "Hugging Face daily paper entry for DrawMotion: Generating 3D Human Motions by Freehand Drawing (2605.20955).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20955", "paper": "https://arxiv.org/abs/2605.20955", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20179", "mode": "signal", "kind": "paper", "title": "TIDE: Efficient and Lossless MoE Diffusion LLM Inference with I/O-aware Expert Offload", "summary": "Hugging Face daily paper entry for TIDE: Efficient and Lossless MoE Diffusion LLM Inference with I/O-aware Expert Offload (2605.20179).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20179", "paper": "https://arxiv.org/abs/2605.20179", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.20158", "mode": "signal", "kind": "paper", "title": "Rethinking Visual Attribution for Chest X-ray Reasoning in Large Vision Language Models", "summary": "Hugging Face daily paper entry for Rethinking Visual Attribution for Chest X-ray Reasoning in Large Vision Language Models (2605.20158).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.20158", "paper": "https://arxiv.org/abs/2605.20158", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.19008", "mode": "signal", "kind": "paper", "title": "Learn-by-Wire Training Control Governance: Bounded Autonomous Training Under Stress for Stability and Efficiency", "summary": "Hugging Face daily paper entry for Learn-by-Wire Training Control Governance: Bounded Autonomous Training Under Stress for Stability and Efficiency (2605.19008).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19008", "paper": "https://arxiv.org/abs/2605.19008", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.18329", "mode": "signal", "kind": "paper", "title": "Lost in the Folds: When Cross-Validation Is Not a Deep Ensemble for Uncertainty Estimation", "summary": "Hugging Face daily paper entry for Lost in the Folds: When Cross-Validation Is Not a Deep Ensemble for Uncertainty Estimation (2605.18329).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18329", "paper": "https://arxiv.org/abs/2605.18329", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17916", "mode": "signal", "kind": "paper", "title": "PanoWorld: A Generative Spatial World Model for Consistent Whole-House Panorama Synthesis", "summary": "Hugging Face daily paper entry for PanoWorld: A Generative Spatial World Model for Consistent Whole-House Panorama Synthesis (2605.17916).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17916", "paper": "https://arxiv.org/abs/2605.17916", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17110", "mode": "signal", "kind": "paper", "title": "Capturing LLM Capabilities via Evidence-Calibrated Query Clustering", "summary": "Hugging Face daily paper entry for Capturing LLM Capabilities via Evidence-Calibrated Query Clustering (2605.17110).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17110", "paper": "https://arxiv.org/abs/2605.17110", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17109", "mode": "signal", "kind": "paper", "title": "DynMuon: A Dynamic Spectral Shaping View of Muon", "summary": "Hugging Face daily paper entry for DynMuon: A Dynamic Spectral Shaping View of Muon (2605.17109).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17109", "paper": "https://arxiv.org/abs/2605.17109", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2602.07892", "mode": "signal", "kind": "paper", "title": "Safety Alignment as Continual Learning: Mitigating the Alignment Tax via Orthogonal Gradient Projection", "summary": "Hugging Face daily paper entry for Safety Alignment as Continual Learning: Mitigating the Alignment Tax via Orthogonal Gradient Projection (2602.07892).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-21", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2602.07892", "paper": "https://arxiv.org/abs/2602.07892", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-21", "sourceUpdatedAt": "2026-05-21", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.11609", "mode": "signal", "kind": "paper", "title": "Anti-Self-Distillation for Reasoning RL via Pointwise Mutual Information", "summary": "Anti-Self-Distillation reverses the direction of knowledge transfer in self-distillation to improve math reasoning efficiency and accuracy. On-policy self-distillation , where a student is pulled toward a copy of itself conditioned on privileged context (e....", "popularity": { "value": 195, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.11609", "paper": "https://arxiv.org/abs/2605.11609", "repo": "https://github.com/FloyedShen/AntiSD", "docs": "https://arxiv.org/pdf/2605.11609" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20025", "mode": "signal", "kind": "paper", "title": "AutoResearchClaw: Self-Reinforcing Autonomous Research with Human-AI Collaboration", "summary": "AutoResearchClaw is a multi-agent autonomous research system that improves scientific discovery through structured debate, self-healing execution, verifiable reporting, human collaboration, and evolutionary learning, outperforming previous systems on a benc...", "popularity": { "value": 185, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20025", "paper": "https://arxiv.org/abs/2605.20025", "repo": "https://github.com/aiming-lab/AutoResearchClaw", "docs": "https://arxiv.org/pdf/2605.20025" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16403", "mode": "signal", "kind": "paper", "title": "When Vision Speaks for Sound", "summary": "Hugging Face daily paper entry for When Vision Speaks for Sound (2605.16403).", "popularity": { "value": 149, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.16403", "paper": "https://arxiv.org/abs/2605.16403", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14236", "mode": "signal", "kind": "paper", "title": "Active Learners as Efficient PRP Rerankers", "summary": "Hugging Face daily paper entry for Active Learners as Efficient PRP Rerankers (2605.14236).", "popularity": { "value": 98, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "embeddings", "tags": [ "embeddings", "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.14236", "paper": "https://arxiv.org/abs/2605.14236", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19769", "mode": "signal", "kind": "paper", "title": "OpenComputer: Verifiable Software Worlds for Computer-Use Agents", "summary": "Hugging Face daily paper entry for OpenComputer: Verifiable Software Worlds for Computer-Use Agents (2605.19769).", "popularity": { "value": 81, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19769", "paper": "https://arxiv.org/abs/2605.19769", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19577", "mode": "signal", "kind": "paper", "title": "GoLongRL: Capability-Oriented Long Context Reinforcement Learning with Multitask Alignment", "summary": "Hugging Face daily paper entry for GoLongRL: Capability-Oriented Long Context Reinforcement Learning with Multitask Alignment (2605.19577).", "popularity": { "value": 58, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19577", "paper": "https://arxiv.org/abs/2605.19577", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15529", "mode": "signal", "kind": "paper", "title": "Process Rewards with Learned Reliability", "summary": "Hugging Face daily paper entry for Process Rewards with Learned Reliability (2605.15529).", "popularity": { "value": 53, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15529", "paper": "https://arxiv.org/abs/2605.15529", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18703", "mode": "signal", "kind": "paper", "title": "EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL", "summary": "Hugging Face daily paper entry for EnvFactory: Scaling Tool-Use Agents via Executable Environments Synthesis and Robust RL (2605.18703).", "popularity": { "value": 50, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18703", "paper": "https://arxiv.org/abs/2605.18703", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17734", "mode": "signal", "kind": "paper", "title": "Harnessing LLM Agents with Skill Programs", "summary": "Hugging Face daily paper entry for Harnessing LLM Agents with Skill Programs (2605.17734).", "popularity": { "value": 36, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17734", "paper": "https://arxiv.org/abs/2605.17734", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19995", "mode": "signal", "kind": "paper", "title": "CogOmniControl: Reasoning-Driven Controllable Video Generation via Creative Intent Cognition", "summary": "Hugging Face daily paper entry for CogOmniControl: Reasoning-Driven Controllable Video Generation via Creative Intent Cognition (2605.19995).", "popularity": { "value": 34, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.19995", "paper": "https://arxiv.org/abs/2605.19995", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18748", "mode": "signal", "kind": "paper", "title": "Aurora: Unified Video Editing with a Tool-Using Agent", "summary": "Hugging Face daily paper entry for Aurora: Unified Video Editing with a Tool-Using Agent (2605.18748).", "popularity": { "value": 29, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18748", "paper": "https://arxiv.org/abs/2605.18748", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18984", "mode": "signal", "kind": "paper", "title": "Artifact-Bench: Evaluating MLLMs on Detecting and Assessing the Artifacts of AI-Generated Videos", "summary": "Hugging Face daily paper entry for Artifact-Bench: Evaluating MLLMs on Detecting and Assessing the Artifacts of AI-Generated Videos (2605.18984).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18984", "paper": "https://arxiv.org/abs/2605.18984", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20087", "mode": "signal", "kind": "paper", "title": "ThoughtTrace: Understanding User Thoughts in Real-World LLM Interactions", "summary": "Hugging Face daily paper entry for ThoughtTrace: Understanding User Thoughts in Real-World LLM Interactions (2605.20087).", "popularity": { "value": 18, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20087", "paper": "https://arxiv.org/abs/2605.20087", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18758", "mode": "signal", "kind": "paper", "title": "OmniGUI: Benchmarking GUI Agents in Omni-Modal Smartphone Environments", "summary": "Hugging Face daily paper entry for OmniGUI: Benchmarking GUI Agents in Omni-Modal Smartphone Environments (2605.18758).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18758", "paper": "https://arxiv.org/abs/2605.18758", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20183", "mode": "signal", "kind": "paper", "title": "MSAVBench: Towards Comprehensive and Reliable Evaluation of Multi-Shot Audio-Video Generation", "summary": "Hugging Face daily paper entry for MSAVBench: Towards Comprehensive and Reliable Evaluation of Multi-Shot Audio-Video Generation (2605.20183).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20183", "paper": "https://arxiv.org/abs/2605.20183", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19436", "mode": "signal", "kind": "paper", "title": "CEPO: RLVR Self-Distillation using Contrastive Evidence Policy Optimization", "summary": "Hugging Face daily paper entry for CEPO: RLVR Self-Distillation using Contrastive Evidence Policy Optimization (2605.19436).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19436", "paper": "https://arxiv.org/abs/2605.19436", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17829", "mode": "signal", "kind": "paper", "title": "Interactive Evaluation Requires a Design Science", "summary": "Hugging Face daily paper entry for Interactive Evaluation Requires a Design Science (2605.17829).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17829", "paper": "https://arxiv.org/abs/2605.17829", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18101", "mode": "signal", "kind": "paper", "title": "SENSE: Satellite-based ENergy Synthesis for Sustainable Environment", "summary": "Hugging Face daily paper entry for SENSE: Satellite-based ENergy Synthesis for Sustainable Environment (2605.18101).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18101", "paper": "https://arxiv.org/abs/2605.18101", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20147", "mode": "signal", "kind": "paper", "title": "PixVerve: Advancing Native UHR Image Generation to 100MP with a Large-Scale High-Quality Dataset", "summary": "Hugging Face daily paper entry for PixVerve: Advancing Native UHR Image Generation to 100MP with a Large-Scale High-Quality Dataset (2605.20147).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research", "vision" ], "links": { "primary": "https://huggingface.co/papers/2605.20147", "paper": "https://arxiv.org/abs/2605.20147", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18714", "mode": "signal", "kind": "paper", "title": "Semantic Generative Tuning for Unified Multimodal Models", "summary": "Hugging Face daily paper entry for Semantic Generative Tuning for Unified Multimodal Models (2605.18714).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18714", "paper": "https://arxiv.org/abs/2605.18714", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15458", "mode": "signal", "kind": "paper", "title": "Video Models Can Reason with Verifiable Rewards", "summary": "Hugging Face daily paper entry for Video Models Can Reason with Verifiable Rewards (2605.15458).", "popularity": { "value": 11, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.15458", "paper": "https://arxiv.org/abs/2605.15458", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19786", "mode": "signal", "kind": "paper", "title": "Fast 4D Mesh Generation by Spatio-Temporal Attention Chains", "summary": "Hugging Face daily paper entry for Fast 4D Mesh Generation by Spatio-Temporal Attention Chains (2605.19786).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19786", "paper": "https://arxiv.org/abs/2605.19786", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18263", "mode": "signal", "kind": "paper", "title": "RT-Splatting: Joint Reflection-Transmission Modeling with Gaussian Splatting", "summary": "Hugging Face daily paper entry for RT-Splatting: Joint Reflection-Transmission Modeling with Gaussian Splatting (2605.18263).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18263", "paper": "https://arxiv.org/abs/2605.18263", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18855", "mode": "signal", "kind": "paper", "title": "Delta Attention Residuals", "summary": "Hugging Face daily paper entry for Delta Attention Residuals (2605.18855).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18855", "paper": "https://arxiv.org/abs/2605.18855", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.09640", "mode": "signal", "kind": "paper", "title": "Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning", "summary": "Hugging Face daily paper entry for Overcoming Catastrophic Forgetting in Visual Continual Learning with Reinforcement Fine-Tuning (2605.09640).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.09640", "paper": "https://arxiv.org/abs/2605.09640", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20150", "mode": "signal", "kind": "paper", "title": "TideGS: Scalable Training of Over One Billion 3D Gaussian Splatting Primitives via Out-of-Core Optimization", "summary": "Hugging Face daily paper entry for TideGS: Scalable Training of Over One Billion 3D Gaussian Splatting Primitives via Out-of-Core Optimization (2605.20150).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20150", "paper": "https://arxiv.org/abs/2605.20150", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20104", "mode": "signal", "kind": "paper", "title": "Draft Less, Retrieve More: Hybrid Tree Construction for Speculative Decoding", "summary": "Hugging Face daily paper entry for Draft Less, Retrieve More: Hybrid Tree Construction for Speculative Decoding (2605.20104).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20104", "paper": "https://arxiv.org/abs/2605.20104", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19932", "mode": "signal", "kind": "paper", "title": "PEEK: Context Map as an Orientation Cache for Long-Context LLM Agents", "summary": "Hugging Face daily paper entry for PEEK: Context Map as an Orientation Cache for Long-Context LLM Agents (2605.19932).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19932", "paper": "https://arxiv.org/abs/2605.19932", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18827", "mode": "signal", "kind": "paper", "title": "Code-Guided Reasoning for Small Language Models: Evaluating Executable MCQA Scaffolds", "summary": "Hugging Face daily paper entry for Code-Guided Reasoning for Small Language Models: Evaluating Executable MCQA Scaffolds (2605.18827).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18827", "paper": "https://arxiv.org/abs/2605.18827", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20164", "mode": "signal", "kind": "paper", "title": "Not Every Rubric Teaches Equally: Policy-Aware Rubric Rewards for RLVR", "summary": "Hugging Face daily paper entry for Not Every Rubric Teaches Equally: Policy-Aware Rubric Rewards for RLVR (2605.20164).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20164", "paper": "https://arxiv.org/abs/2605.20164", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19633", "mode": "signal", "kind": "paper", "title": "optimize_anything: A Universal API for Optimizing any Text Parameter", "summary": "Hugging Face daily paper entry for optimize_anything: A Universal API for Optimizing any Text Parameter (2605.19633).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19633", "paper": "https://arxiv.org/abs/2605.19633", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19305", "mode": "signal", "kind": "paper", "title": "Matérn Noise for Triangulation-Agnostic Flow Matching on Meshes", "summary": "Hugging Face daily paper entry for Matérn Noise for Triangulation-Agnostic Flow Matching on Meshes (2605.19305).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19305", "paper": "https://arxiv.org/abs/2605.19305", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18226", "mode": "signal", "kind": "paper", "title": "Context Memorization for Efficient Long Context Generation", "summary": "Hugging Face daily paper entry for Context Memorization for Efficient Long Context Generation (2605.18226).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.18226", "paper": "https://arxiv.org/abs/2605.18226", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.09789", "mode": "signal", "kind": "paper", "title": "Zero-Shot Sim-to-Real Robot Learning: A Dexterous Manipulation Study on Reactive Catching", "summary": "Hugging Face daily paper entry for Zero-Shot Sim-to-Real Robot Learning: A Dexterous Manipulation Study on Reactive Catching (2605.09789).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.09789", "paper": "https://arxiv.org/abs/2605.09789", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20035", "mode": "signal", "kind": "paper", "title": "Stage-adaptive Token Selection for Efficient Omni-modal LLMs", "summary": "Hugging Face daily paper entry for Stage-adaptive Token Selection for Efficient Omni-modal LLMs (2605.20035).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.20035", "paper": "https://arxiv.org/abs/2605.20035", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.19908", "mode": "signal", "kind": "paper", "title": "Where Does Authorship Signal Emerge in Encoder-Based Language Models?", "summary": "Hugging Face daily paper entry for Where Does Authorship Signal Emerge in Encoder-Based Language Models? (2605.19908).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19908", "paper": "https://arxiv.org/abs/2605.19908", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18746", "mode": "signal", "kind": "paper", "title": "ESI-Bench: Towards Embodied Spatial Intelligence that Closes the Perception-Action Loop", "summary": "Hugging Face daily paper entry for ESI-Bench: Towards Embodied Spatial Intelligence that Closes the Perception-Action Loop (2605.18746).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18746", "paper": "https://arxiv.org/abs/2605.18746", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16003", "mode": "signal", "kind": "paper", "title": "Echo-Forcing: A Scene Memory Framework for Interactive Long Video Generation", "summary": "Hugging Face daily paper entry for Echo-Forcing: A Scene Memory Framework for Interactive Long Video Generation (2605.16003).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.16003", "paper": "https://arxiv.org/abs/2605.16003", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.08472", "mode": "signal", "kind": "paper", "title": "Mid-Training with Self-Generated Data Improves Reinforcement Learning in Language Models", "summary": "Hugging Face daily paper entry for Mid-Training with Self-Generated Data Improves Reinforcement Learning in Language Models (2605.08472).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.08472", "paper": "https://arxiv.org/abs/2605.08472", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.20075", "mode": "signal", "kind": "paper", "title": "CopT: Contrastive On-Policy Thinking with Continuous Spaces for General and Agentic Reasoning", "summary": "Hugging Face daily paper entry for CopT: Contrastive On-Policy Thinking with Continuous Spaces for General and Agentic Reasoning (2605.20075).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.20075", "paper": "https://arxiv.org/abs/2605.20075", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.19516", "mode": "signal", "kind": "paper", "title": "Base Models Look Human To AI Detectors", "summary": "Hugging Face daily paper entry for Base Models Look Human To AI Detectors (2605.19516).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19516", "paper": "https://arxiv.org/abs/2605.19516", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.19147", "mode": "signal", "kind": "paper", "title": "Be Kind, Rewrite: Benign Projections via Rewriting Defend Against LLM Data Poisoning Attacks", "summary": "Hugging Face daily paper entry for Be Kind, Rewrite: Benign Projections via Rewriting Defend Against LLM Data Poisoning Attacks (2605.19147).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19147", "paper": "https://arxiv.org/abs/2605.19147", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.19014", "mode": "signal", "kind": "paper", "title": "SAGA: A Sequence-Adaptive Generative Architecture for Multi-Horizon Probabilistic Forecasting with Adaptive Temporal Conformal Prediction", "summary": "Hugging Face daily paper entry for SAGA: A Sequence-Adaptive Generative Architecture for Multi-Horizon Probabilistic Forecasting with Adaptive Temporal Conformal Prediction (2605.19014).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.19014", "paper": "https://arxiv.org/abs/2605.19014", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.18646", "mode": "signal", "kind": "paper", "title": "Language-Switching Triggers Take a Latent Detour Through Language Models", "summary": "Hugging Face daily paper entry for Language-Switching Triggers Take a Latent Detour Through Language Models (2605.18646).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18646", "paper": "https://arxiv.org/abs/2605.18646", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17909", "mode": "signal", "kind": "paper", "title": "Ethical Hyper-Velocity (EHV): A Provably Deterministic Governance-Aware JIT Compiler Architecture for Agentic Systems", "summary": "Hugging Face daily paper entry for Ethical Hyper-Velocity (EHV): A Provably Deterministic Governance-Aware JIT Compiler Architecture for Agentic Systems (2605.17909).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17909", "paper": "https://arxiv.org/abs/2605.17909", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17659", "mode": "signal", "kind": "paper", "title": "Bug or Feature^2: Weight Drift, Activation Sparsity, and Spikes", "summary": "Hugging Face daily paper entry for Bug or Feature^2: Weight Drift, Activation Sparsity, and Spikes (2605.17659).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17659", "paper": "https://arxiv.org/abs/2605.17659", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17360", "mode": "signal", "kind": "paper", "title": "Omni-DuplexEval: Evaluating Real-time Duplex Omni-modal Interaction", "summary": "Hugging Face daily paper entry for Omni-DuplexEval: Evaluating Real-time Duplex Omni-modal Interaction (2605.17360).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17360", "paper": "https://arxiv.org/abs/2605.17360", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17076", "mode": "signal", "kind": "paper", "title": "S-Bus: Automatic Read-Set Reconstruction for Multi-Agent LLM State Coordination", "summary": "Hugging Face daily paper entry for S-Bus: Automatic Read-Set Reconstruction for Multi-Agent LLM State Coordination (2605.17076).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17076", "paper": "https://arxiv.org/abs/2605.17076", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17026", "mode": "signal", "kind": "paper", "title": "Why Do Reasoning Models Lose Coverage? The Role of Data and Forks in the Road", "summary": "Hugging Face daily paper entry for Why Do Reasoning Models Lose Coverage? The Role of Data and Forks in the Road (2605.17026).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Retrieval research is directly relevant to private knowledge assistants, codebase QA, and grounded AI systems.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "rag", "tags": [ "huggingface-papers", "paper", "rag", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17026", "paper": "https://arxiv.org/abs/2605.17026", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.15514", "mode": "signal", "kind": "paper", "title": "RoPE Distinguishes Neither Positions Nor Tokens in Long Contexts, Provably", "summary": "Hugging Face daily paper entry for RoPE Distinguishes Neither Positions Nor Tokens in Long Contexts, Provably (2605.15514).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15514", "paper": "https://arxiv.org/abs/2605.15514", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.14842", "mode": "signal", "kind": "paper", "title": "Editor's Choice: Evaluating Abstract Intent in Image Editing through Atomic Entity Analysis", "summary": "Hugging Face daily paper entry for Editor's Choice: Evaluating Abstract Intent in Image Editing through Atomic Entity Analysis (2605.14842).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.14842", "paper": "https://arxiv.org/abs/2605.14842", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.12623", "mode": "signal", "kind": "paper", "title": "DocAtlas: Multilingual Document Understanding Across 80+ Languages", "summary": "Hugging Face daily paper entry for DocAtlas: Multilingual Document Understanding Across 80+ Languages (2605.12623).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.12623", "paper": "https://arxiv.org/abs/2605.12623", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.08586", "mode": "signal", "kind": "paper", "title": "Computer Science Conferences Should Require Nonrepudiable Experimental Results", "summary": "Hugging Face daily paper entry for Computer Science Conferences Should Require Nonrepudiable Experimental Results (2605.08586).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-20", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.08586", "paper": "https://arxiv.org/abs/2605.08586", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-20", "sourceUpdatedAt": "2026-05-20", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "hf-dataset:1111xxx/zoengjyutgaai", "mode": "catalog", "kind": "dataset", "title": "1111xxx/zoengjyutgaai", "summary": "Hugging Face dataset with 55950 downloads.", "popularity": { "value": 55950, "label": "downloads" }, "whyItMatters": "Useful to archive because it is a source-listed dataset with measurable Hugging Face usage.", "sourceId": "huggingface-datasets", "sourceName": "Hugging Face Datasets", "sourceUrl": "https://huggingface.co/api/datasets?sort=downloads&direction=-1&limit=30", "publishedAt": "2026-05-19", "category": "datasets", "subcategory": "datasets", "tags": [ "art", "audio", "cantonese", "datasets", "language:yue", "license:cc0-1.0", "modality:audio", "region:us" ], "links": { "primary": "https://huggingface.co/datasets/1111xxx/zoengjyutgaai", "paper": "", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": false, "license": "cc0-1.0", "status": "active", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "hf-dataset:Wan-fen52/MAVOS-DD", "mode": "catalog", "kind": "dataset", "title": "Wan-fen52/MAVOS-DD", "summary": "Hugging Face dataset with 52911 downloads.", "popularity": { "value": 52911, "label": "downloads" }, "whyItMatters": "Useful to archive because it is a source-listed dataset with measurable Hugging Face usage.", "sourceId": "huggingface-datasets", "sourceName": "Hugging Face Datasets", "sourceUrl": "https://huggingface.co/api/datasets?sort=downloads&direction=-1&limit=30", "publishedAt": "2026-05-19", "category": "datasets", "subcategory": "datasets", "tags": [ "arxiv:2505.11109", "datasets", "language:ar", "language:de", "language:en", "language:es", "language:hi", "language:ro" ], "links": { "primary": "https://huggingface.co/datasets/Wan-fen52/MAVOS-DD", "paper": "", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": false, "license": "unknown", "status": "active", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "paper:arxiv:2605.18747", "mode": "signal", "kind": "paper", "title": "Code as Agent Harness", "summary": "Large language models are increasingly used as operational substrates for agent reasoning and execution in agentic systems, with code serving as a unified infrastructure layer across multiple domains and applications. Recent large language models (LLMs) hav...", "popularity": { "value": 211, "label": "upvotes" }, "whyItMatters": "This paper includes implementation links, making it easier for builders to test the idea instead of only reading the method.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18747", "paper": "https://arxiv.org/abs/2605.18747", "repo": "https://github.com/YennNing/Awesome-Code-as-Agent-Harness-Papers", "docs": "https://arxiv.org/pdf/2605.18747" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18401", "mode": "signal", "kind": "paper", "title": "SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution", "summary": "Hugging Face daily paper entry for SkillsVote: Lifecycle Governance of Agent Skills from Collection, Recommendation to Evolution (2605.18401).", "popularity": { "value": 126, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18401", "paper": "https://arxiv.org/abs/2605.18401", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18739", "mode": "signal", "kind": "paper", "title": "LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation", "summary": "Hugging Face daily paper entry for LongLive-2.0: An NVFP4 Parallel Infrastructure for Long Video Generation (2605.18739).", "popularity": { "value": 112, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.18739", "paper": "https://arxiv.org/abs/2605.18739", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18678", "mode": "signal", "kind": "paper", "title": "Lance: Unified Multimodal Modeling by Multi-Task Synergy", "summary": "Hugging Face daily paper entry for Lance: Unified Multimodal Modeling by Multi-Task Synergy (2605.18678).", "popularity": { "value": 78, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18678", "paper": "https://arxiv.org/abs/2605.18678", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18661", "mode": "signal", "kind": "paper", "title": "AI for Auto-Research: Roadmap & User Guide", "summary": "Hugging Face daily paper entry for AI for Auto-Research: Roadmap & User Guide (2605.18661).", "popularity": { "value": 67, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18661", "paper": "https://arxiv.org/abs/2605.18661", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17757", "mode": "signal", "kind": "paper", "title": "OSCAR: Offline Spectral Covariance-Aware Rotation for 2-bit KV Cache Quantization", "summary": "Hugging Face daily paper entry for OSCAR: Offline Spectral Covariance-Aware Rotation for 2-bit KV Cache Quantization (2605.17757).", "popularity": { "value": 63, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17757", "paper": "https://arxiv.org/abs/2605.17757", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16679", "mode": "signal", "kind": "paper", "title": "CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows?", "summary": "Hugging Face daily paper entry for CHI-Bench: Can AI Agents Automate End-to-End, Long-Horizon, Policy-Rich Healthcare Workflows? (2605.16679).", "popularity": { "value": 53, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16679", "paper": "https://arxiv.org/abs/2605.16679", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18451", "mode": "signal", "kind": "paper", "title": "Code-as-Room: Generating 3D Rooms from Top-Down View Images via Agentic Code Synthesis", "summary": "Hugging Face daily paper entry for Code-as-Room: Generating 3D Rooms from Top-Down View Images via Agentic Code Synthesis (2605.18451).", "popularity": { "value": 41, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18451", "paper": "https://arxiv.org/abs/2605.18451", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14278", "mode": "signal", "kind": "paper", "title": "KVPO: ODE-Native GRPO for Autoregressive Video Alignment via KV Semantic Exploration", "summary": "Hugging Face daily paper entry for KVPO: ODE-Native GRPO for Autoregressive Video Alignment via KV Semantic Exploration (2605.14278).", "popularity": { "value": 37, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "fine-tuning", "tags": [ "fine-tuning", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.14278", "paper": "https://arxiv.org/abs/2605.14278", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17283", "mode": "signal", "kind": "paper", "title": "OProver: A Unified Framework for Agentic Formal Theorem Proving", "summary": "Hugging Face daily paper entry for OProver: A Unified Framework for Agentic Formal Theorem Proving (2605.17283).", "popularity": { "value": 31, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17283", "paper": "https://arxiv.org/abs/2605.17283", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18643", "mode": "signal", "kind": "paper", "title": "Post-Trained MoE Can Skip Half Experts via Self-Distillation", "summary": "Hugging Face daily paper entry for Post-Trained MoE Can Skip Half Experts via Self-Distillation (2605.18643).", "popularity": { "value": 30, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18643", "paper": "https://arxiv.org/abs/2605.18643", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16079", "mode": "signal", "kind": "paper", "title": "VideoSeeker: Incentivizing Instance-level Video Understanding via Native Agentic Tool Invocation", "summary": "Hugging Face daily paper entry for VideoSeeker: Incentivizing Instance-level Video Understanding via Native Agentic Tool Invocation (2605.16079).", "popularity": { "value": 28, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16079", "paper": "https://arxiv.org/abs/2605.16079", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17260", "mode": "signal", "kind": "paper", "title": "LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMs", "summary": "Hugging Face daily paper entry for LiteFrame: Efficient Vision Encoders Unlock Frame Scaling in Video LLMs (2605.17260).", "popularity": { "value": 25, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "small-local", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.17260", "paper": "https://arxiv.org/abs/2605.17260", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17672", "mode": "signal", "kind": "paper", "title": "Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models", "summary": "Hugging Face daily paper entry for Stop When Reasoning Converges: Semantic-Preserving Early Exit for Reasoning Models (2605.17672).", "popularity": { "value": 22, "label": "upvotes" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17672", "paper": "https://arxiv.org/abs/2605.17672", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15572", "mode": "signal", "kind": "paper", "title": "Measuring Maximum Activations in Open Large Language Models", "summary": "Hugging Face daily paper entry for Measuring Maximum Activations in Open Large Language Models (2605.15572).", "popularity": { "value": 18, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15572", "paper": "https://arxiv.org/abs/2605.15572", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14589", "mode": "signal", "kind": "paper", "title": "EndPrompt: Efficient Long-Context Extension via Terminal Anchoring", "summary": "Hugging Face daily paper entry for EndPrompt: Efficient Long-Context Extension via Terminal Anchoring (2605.14589).", "popularity": { "value": 17, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.14589", "paper": "https://arxiv.org/abs/2605.14589", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15565", "mode": "signal", "kind": "paper", "title": "AstraFlow: Dataflow-Oriented Reinforcement Learning for Agentic LLMs", "summary": "Hugging Face daily paper entry for AstraFlow: Dataflow-Oriented Reinforcement Learning for Agentic LLMs (2605.15565).", "popularity": { "value": 16, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15565", "paper": "https://arxiv.org/abs/2605.15565", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18287", "mode": "signal", "kind": "paper", "title": "StableVLA: Towards Robust Vision-Language-Action Models without Extra Data", "summary": "Hugging Face daily paper entry for StableVLA: Towards Robust Vision-Language-Action Models without Extra Data (2605.18287).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18287", "paper": "https://arxiv.org/abs/2605.18287", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14038", "mode": "signal", "kind": "paper", "title": "Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use", "summary": "Hugging Face daily paper entry for Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use (2605.14038).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.14038", "paper": "https://arxiv.org/abs/2605.14038", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.12290", "mode": "signal", "kind": "paper", "title": "Targeted Neuron Modulation via Contrastive Pair Search", "summary": "Hugging Face daily paper entry for Targeted Neuron Modulation via Contrastive Pair Search (2605.12290).", "popularity": { "value": 15, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.12290", "paper": "https://arxiv.org/abs/2605.12290", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.14368", "mode": "signal", "kind": "paper", "title": "Where Should Diffusion Enter a Language Model? Geometry-Guided Hidden-State Replacement", "summary": "Hugging Face daily paper entry for Where Should Diffusion Enter a Language Model? Geometry-Guided Hidden-State Replacement (2605.14368).", "popularity": { "value": 14, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.14368", "paper": "https://arxiv.org/abs/2605.14368", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16839", "mode": "signal", "kind": "paper", "title": "CompactAttention: Accelerating Chunked Prefill with Block-Union KV Selection", "summary": "Hugging Face daily paper entry for CompactAttention: Accelerating Chunked Prefill with Block-Union KV Selection (2605.16839).", "popularity": { "value": 13, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16839", "paper": "https://arxiv.org/abs/2605.16839", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17242", "mode": "signal", "kind": "paper", "title": "From Runnable to Shippable: Multi-Agent Test-Driven Development for Generating Full-Stack Web Applications from Requirements", "summary": "Hugging Face daily paper entry for From Runnable to Shippable: Multi-Agent Test-Driven Development for Generating Full-Stack Web Applications from Requirements (2605.17242).", "popularity": { "value": 12, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17242", "paper": "https://arxiv.org/abs/2605.17242", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18749", "mode": "signal", "kind": "paper", "title": "WavFlow: Audio Generation in Waveform Space", "summary": "Hugging Face daily paper entry for WavFlow: Audio Generation in Waveform Space (2605.18749).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18749", "paper": "https://arxiv.org/abs/2605.18749", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16893", "mode": "signal", "kind": "paper", "title": "NGM: A Plug-and-Play Training-Free Memory Module for LLMs", "summary": "Hugging Face daily paper entry for NGM: A Plug-and-Play Training-Free Memory Module for LLMs (2605.16893).", "popularity": { "value": 10, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16893", "paper": "https://arxiv.org/abs/2605.16893", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16909", "mode": "signal", "kind": "paper", "title": "TOBench: A Task-Oriented Omni-Modal Benchmark for Real-World Tool-Using Agents", "summary": "Hugging Face daily paper entry for TOBench: A Task-Oriented Omni-Modal Benchmark for Real-World Tool-Using Agents (2605.16909).", "popularity": { "value": 9, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16909", "paper": "https://arxiv.org/abs/2605.16909", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18652", "mode": "signal", "kind": "paper", "title": "MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents", "summary": "Hugging Face daily paper entry for MementoGUI: Learning Agentic Multimodal Memory Control for Long-Horizon GUI Agents (2605.18652).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18652", "paper": "https://arxiv.org/abs/2605.18652", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.15482", "mode": "signal", "kind": "paper", "title": "FINESSE-Bench: A Hierarchical Benchmark Suite for Financial Domain Knowledge and Technical Analysis in Large Language Models", "summary": "Hugging Face daily paper entry for FINESSE-Bench: A Hierarchical Benchmark Suite for Financial Domain Knowledge and Technical Analysis in Large Language Models (2605.15482).", "popularity": { "value": 8, "label": "upvotes" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15482", "paper": "https://arxiv.org/abs/2605.15482", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18727", "mode": "signal", "kind": "paper", "title": "DexHoldem: Playing Texas Hold'em with Dexterous Embodied System", "summary": "Hugging Face daily paper entry for DexHoldem: Playing Texas Hold'em with Dexterous Embodied System (2605.18727).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18727", "paper": "https://arxiv.org/abs/2605.18727", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17933", "mode": "signal", "kind": "paper", "title": "AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents", "summary": "Hugging Face daily paper entry for AtlasVA: Self-Evolving Visual Skill Memory for Teacher-Free VLM Agents (2605.17933).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17933", "paper": "https://arxiv.org/abs/2605.17933", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.16865", "mode": "signal", "kind": "paper", "title": "MixSD: Mixed Contextual Self-Distillation for Knowledge Injection", "summary": "Hugging Face daily paper entry for MixSD: Mixed Contextual Self-Distillation for Knowledge Injection (2605.16865).", "popularity": { "value": 7, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "small-local", "tags": [ "huggingface-papers", "paper", "research", "small-local" ], "links": { "primary": "https://huggingface.co/papers/2605.16865", "paper": "https://arxiv.org/abs/2605.16865", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18601", "mode": "signal", "kind": "paper", "title": "Incantation: Natural Language as the Action Interface for Multi-Entity Video World Models", "summary": "Hugging Face daily paper entry for Incantation: Natural Language as the Action Interface for Multi-Entity Video World Models (2605.18601).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "video-generation", "tags": [ "huggingface-papers", "paper", "research", "video-generation" ], "links": { "primary": "https://huggingface.co/papers/2605.18601", "paper": "https://arxiv.org/abs/2605.18601", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17698", "mode": "signal", "kind": "paper", "title": "Agent Bazaar: Enabling Economic Alignment in Multi-Agent Marketplaces", "summary": "Hugging Face daily paper entry for Agent Bazaar: Enabling Economic Alignment in Multi-Agent Marketplaces (2605.17698).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17698", "paper": "https://arxiv.org/abs/2605.17698", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17199", "mode": "signal", "kind": "paper", "title": "Geometric Phase Transition Enables Extreme Hippocampal Memory Capacity", "summary": "Hugging Face daily paper entry for Geometric Phase Transition Enables Extreme Hippocampal Memory Capacity (2605.17199).", "popularity": { "value": 6, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17199", "paper": "https://arxiv.org/abs/2605.17199", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18719", "mode": "signal", "kind": "paper", "title": "SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training", "summary": "Hugging Face daily paper entry for SafeDiffusion-R1: Online Reward Steering for Safe Diffusion Post-Training (2605.18719).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "image-generation", "tags": [ "huggingface-papers", "image-generation", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18719", "paper": "https://arxiv.org/abs/2605.18719", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.17894", "mode": "signal", "kind": "paper", "title": "Evaluating Cognitive Age Alignment in Interactive AI Agents", "summary": "Hugging Face daily paper entry for Evaluating Cognitive Age Alignment in Interactive AI Agents (2605.17894).", "popularity": { "value": 5, "label": "upvotes" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17894", "paper": "https://arxiv.org/abs/2605.17894", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-03T05:03:47+00:00" }, { "id": "paper:arxiv:2605.18743", "mode": "signal", "kind": "paper", "title": "Actionable World Representation", "summary": "Hugging Face daily paper entry for Actionable World Representation (2605.18743).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18743", "paper": "https://arxiv.org/abs/2605.18743", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.18630", "mode": "signal", "kind": "paper", "title": "SCICONVBENCH: Benchmarking LLMs on Multi-Turn Clarification for Task Formulation in Computational Science", "summary": "Hugging Face daily paper entry for SCICONVBENCH: Benchmarking LLMs on Multi-Turn Clarification for Task Formulation in Computational Science (2605.18630).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18630", "paper": "https://arxiv.org/abs/2605.18630", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.18549", "mode": "signal", "kind": "paper", "title": "Monitoring the Internal Monologue: Probe Trajectories Reveal Reasoning Dynamics", "summary": "Hugging Face daily paper entry for Monitoring the Internal Monologue: Probe Trajectories Reveal Reasoning Dynamics (2605.18549).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18549", "paper": "https://arxiv.org/abs/2605.18549", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.18106", "mode": "signal", "kind": "paper", "title": "Symmetry-Compatible Principle for Optimizer Design: Embeddings, LM Heads, SwiGLU MLPs, and MoE Routers", "summary": "Hugging Face daily paper entry for Symmetry-Compatible Principle for Optimizer Design: Embeddings, LM Heads, SwiGLU MLPs, and MoE Routers (2605.18106).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "embeddings", "tags": [ "embeddings", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.18106", "paper": "https://arxiv.org/abs/2605.18106", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17842", "mode": "signal", "kind": "paper", "title": "SNLP: Layer-Parallel Inference via Structured Newton Corrections", "summary": "Hugging Face daily paper entry for SNLP: Layer-Parallel Inference via Structured Newton Corrections (2605.17842).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17842", "paper": "https://arxiv.org/abs/2605.17842", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.17278", "mode": "signal", "kind": "paper", "title": "A2RBench: An Automatic Paradigm for Formally Verifiable Abstract Reasoning Benchmark Generation", "summary": "Hugging Face daily paper entry for A2RBench: An Automatic Paradigm for Formally Verifiable Abstract Reasoning Benchmark Generation (2605.17278).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Evaluation papers help builders measure whether models and agents are actually improving.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "evaluation", "tags": [ "evaluation", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.17278", "paper": "https://arxiv.org/abs/2605.17278", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.16882", "mode": "signal", "kind": "paper", "title": "E-PMQ: Expert-Guided Post-Merge Quantization with Merged-Weight Anchoring", "summary": "Hugging Face daily paper entry for E-PMQ: Expert-Guided Post-Merge Quantization with Merged-Weight Anchoring (2605.16882).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Inference improvements can reduce cost and latency for serving or self-hosting models.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "inference", "tags": [ "huggingface-papers", "inference", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16882", "paper": "https://arxiv.org/abs/2605.16882", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.16819", "mode": "signal", "kind": "paper", "title": "AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents", "summary": "Hugging Face daily paper entry for AgentKernelArena: Generalization-Aware Benchmarking of GPU Kernel Optimization Agents (2605.16819).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Agent research can translate into better planning, tool use, and automation patterns for real developer workflows.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "agents", "tags": [ "agents", "huggingface-papers", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16819", "paper": "https://arxiv.org/abs/2605.16819", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.16386", "mode": "signal", "kind": "paper", "title": "Auditing Multimodal LLM Raters: Central Tendency Bias in Clinical Ordinal Scoring", "summary": "Hugging Face daily paper entry for Auditing Multimodal LLM Raters: Central Tendency Bias in Clinical Ordinal Scoring (2605.16386).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.16386", "paper": "https://arxiv.org/abs/2605.16386", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.15764", "mode": "signal", "kind": "paper", "title": "GRASP: Learning to Ground Social Reasoning in Multi-Person Non-Verbal Interactions", "summary": "Hugging Face daily paper entry for GRASP: Learning to Ground Social Reasoning in Multi-Person Non-Verbal Interactions (2605.15764).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15764", "paper": "https://arxiv.org/abs/2605.15764", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2605.15035", "mode": "signal", "kind": "paper", "title": "TopoPrimer: The Missing Topological Context in Forecasting Models", "summary": "Hugging Face daily paper entry for TopoPrimer: The Missing Topological Context in Forecasting Models (2605.15035).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "llm", "tags": [ "huggingface-papers", "llm", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2605.15035", "paper": "https://arxiv.org/abs/2605.15035", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "paper:arxiv:2603.10126", "mode": "signal", "kind": "paper", "title": "AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models", "summary": "Hugging Face daily paper entry for AR-VLA: True Autoregressive Action Expert for Vision-Language-Action Models (2603.10126).", "popularity": { "value": 0, "label": "seen" }, "whyItMatters": "Worth tracking because Hugging Face Daily Papers surfaces research that may become practical models, datasets, or developer tools.", "sourceId": "huggingface-papers", "sourceName": "Hugging Face Papers", "sourceUrl": "https://huggingface.co/papers", "publishedAt": "2026-05-19", "category": "research", "subcategory": "multimodal", "tags": [ "huggingface-papers", "multimodal", "paper", "research" ], "links": { "primary": "https://huggingface.co/papers/2603.10126", "paper": "https://arxiv.org/abs/2603.10126", "repo": "", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "advanced", "maturity": "experimental", "openSource": true, "localFriendly": false, "license": "paper", "status": "new", "createdAt": "2026-05-19", "sourceUpdatedAt": "2026-05-19", "lastSeenAt": "2026-06-01T17:51:08+00:00" }, { "id": "github:lynote-ai/humanize-text", "mode": "catalog", "kind": "repo", "title": "lynote-ai/humanize-text", "summary": "Free open-source AI text humanizer to convert AI-generated content into undetectable, human-like writing. Bypass Turnitin, GPTZero, and all major AI detectors. No sign-up required. Try our unlimited free online tool", "popularity": { "value": 1069, "label": "stars" }, "whyItMatters": "Useful to archive because it is a source-listed GitHub repository with measurable stars.", "sourceId": "github-catalog", "sourceName": "GitHub Catalog", "sourceUrl": "https://api.github.com/search/repositories?q=topic:llm&sort=stars&order=desc&per_page=2", "publishedAt": "2026-06-01", "category": "tools", "subcategory": "developer-tools", "tags": [ "ai-tools", "developer-tools" ], "links": { "primary": "https://github.com/lynote-ai/humanize-text", "paper": "", "repo": "https://github.com/lynote-ai/humanize-text", "docs": "" }, "scores": { "priority": 0, "freshness": 0, "adoption": 0, "breakthrough": 0, "builder": 0, "ease": 0, "local": 0, "hypeRisk": 0 }, "difficulty": "intermediate", "maturity": "active", "openSource": true, "localFriendly": true, "license": "MIT", "status": "active", "createdAt": "2026-05-18", "sourceUpdatedAt": "2026-06-01", "lastSeenAt": "2026-06-04T03:03:53+00:00" }, { "id": "hf-dataset:hrz2001/Maritime_Visual_Tracking_Dataset_MVTD", "mode": "catalog", "kind": "dataset", "title": "hrz2001/Maritime_Visual_Tracking_Dataset_MVTD", "summary": "Hugging Face dataset with 53447 downloads.", "popularity": { "value": 53447, "label": "downloads" }, "whyItMatters": "Useful to archive because it is a source-listed dataset with measurable Hugging Face usage.", "sourceId": "huggingface-datasets", "sourceName": "Hugging Face Datasets", "sourceUrl": "https://huggingface.co/api/datasets?sort=downloads&direction=-1&limit=30", "publishedAt": "2026-05-18", "category": "datasets", "subcategory": "datasets", "tags": [ "arxiv:2506.02866", "datasets", "license:cc0-1.0", "region:us", "size_categories:100k