HF Daily Papers · 2026-10-16

HF Daily Papers 精选（15 篇，按社区票数排序）

  • [63▲] Cortex: Consolidating Context into Parametric Memory — https://arxiv.org/abs/2610.20399
  • [61▲] RuleTrek: Benchmarking Agentic Rule Reasoning in Complex Scenarios — https://arxiv.org/abs/2610.19698
  • [38▲] Many Right Answers, One Training Example: Removing Responder Noise from GRPO for LLM Recommenders — https://arxiv.org/abs/2610.17771
  • [34▲] Apertus 1.5: From Open Text Models to Open Multimodal Reasoners — https://arxiv.org/abs/2610.15542
  • [25▲] CLASHarness: Benchmarking Multi-Agent Conflict in Computer Use via a Controllable and Verifiable Generation Pipeline — https://arxiv.org/abs/2610.19128
  • [20▲] The Thinking Leakage in Reinforcement Learning for LLMs — https://arxiv.org/abs/2610.14270
  • [18▲] Super-Resolution in The Right Latent Space: A Frozen Vision-Foundation Substrate — https://arxiv.org/abs/2610.04781
  • [15▲] LabInstruct: Benchmarking Situated Instructional Video Generation for Lab Procedures — https://arxiv.org/abs/2610.14020
  • [13▲] APEX-Harm: Evaluating Agent Safety in Shared Professional Workspaces — https://arxiv.org/abs/2610.13623
  • [13▲] SwarmRSI: Recursive Self-Improvement through Evolving Scientific Agent Swarm — https://arxiv.org/abs/2610.20431
  • [12▲] ProgAgent: Learning Tool Orchestration through Programmatic Training — https://arxiv.org/abs/2610.14774
  • [9▲] History Replays Itself: Unleashing Long-Context Capability in Linear Attention — https://arxiv.org/abs/2610.15715
  • [9▲] LLM-to-Flow Distillation: Teaching Natural Language Goals to Scientific Generators — https://arxiv.org/abs/2610.17246
  • [9▲] Joint Subspace Sparse Attention for Efficient Long-Context Decoding — https://arxiv.org/abs/2610.19026
  • [8▲] ProbabilityBench: Model Evaluation and Calibration with Exact Ground-Truth Uncertainty — https://arxiv.org/abs/2610.17016