HF Daily Papers · 2026-10-29

HF Daily Papers 精选（15 篇，按社区票数排序）

  • [373▲] Multiscale Policy Optimization for Long-Horizon LLM Agents — https://arxiv.org/abs/2610.35013
  • [345▲] ChronoReplay: One Answer Interface Is Not Enough for Streaming Spatial Intelligence — https://arxiv.org/abs/2610.34283
  • [312▲] The Trinity:Capability-Driven Recursive Self-Improvement for Multimodal Agents — https://arxiv.org/abs/2610.36077
  • [172▲] VibeVideoBench: Evaluating Multimodal Agents on Code-Driven Video Generation — https://arxiv.org/abs/2610.35864
  • [139▲] PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations — https://arxiv.org/abs/2610.37381
  • [124▲] Memory Tax: Dissecting the Net Value of Agent Memory — https://arxiv.org/abs/2610.34073
  • [110▲] Can Test-Time Memory Evolution Outperform Test-Time Training for Scientific Discovery? — https://arxiv.org/abs/2610.32046
  • [105▲] Hy-Nomas: Towards Free-form Interactive World Generation — https://arxiv.org/abs/2610.36729
  • [89▲] Can Discrete Tokens Serve as Semantically Lossless Representations of Visual Signals? — https://arxiv.org/abs/2610.37662
  • [79▲] HABIT-Bench: Evaluating Habit Induction from Longitudinal Weak Evidence in Agent Memory — https://arxiv.org/abs/2610.24520
  • [50▲] CodeCap-Bench: Can Code Serve as a Good Video Caption? — https://arxiv.org/abs/2610.37341
  • [48▲] ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning — https://arxiv.org/abs/2610.37666
  • [30▲] LociArt: A Unified Paradigm for Free-Style Local Video Stylization — https://arxiv.org/abs/2610.36456
  • [29▲] OvisOCR3 Technical Report — https://arxiv.org/abs/2610.33066
  • [27▲] Truncate or Not? A Scaling Study of Looped Transformers with Truncated Backpropagation — https://arxiv.org/abs/2610.26030