HF Daily Papers · 2026-10-30

HF Daily Papers 精选（15 篇，按社区票数排序）

  • [386▲] The Trinity:Capability-Driven Recursive Self-Improvement for Multimodal Agents — https://arxiv.org/abs/2610.36077
  • [380▲] Multiscale Policy Optimization for Long-Horizon LLM Agents — https://arxiv.org/abs/2610.35013
  • [151▲] Memory Tax: Dissecting the Net Value of Agent Memory — https://arxiv.org/abs/2610.34073
  • [149▲] Efficient Streaming Audio-Visual Target Speaker Extraction for Real-World Acoustic Scenes — https://arxiv.org/abs/2610.32749
  • [148▲] PDRVLA: A Vision-Language-Action Model with Physical Dynamics Representations — https://arxiv.org/abs/2610.37381
  • [89▲] VibeVideoBench: Evaluating Multimodal Agents on Code-Driven Video Generation — https://arxiv.org/abs/2610.35864
  • [60▲] Can Discrete Tokens Serve as Semantically Lossless Representations of Visual Signals? — https://arxiv.org/abs/2610.37662
  • [53▲] ThinkWorld: World Imagination Reinforces Chain-of-Dynamics Reasoning — https://arxiv.org/abs/2610.37666
  • [51▲] CodeCap-Bench: Can Code Serve as a Good Video Caption? — https://arxiv.org/abs/2610.37341
  • [44▲] OmniShowcase: Omnidirectional Subject-Consistent Video Generation from Multi-View Images — https://arxiv.org/abs/2610.35684
  • [30▲] Truncate or Not? A Scaling Study of Looped Transformers with Truncated Backpropagation — https://arxiv.org/abs/2610.26030
  • [29▲] OvisOCR3 Technical Report — https://arxiv.org/abs/2610.33066
  • [28▲] Zooming without Zooming v2: From Seeing Details to Connecting Evidence for Generalizable Multimodal Perception — https://arxiv.org/abs/2610.37037
  • [26▲] EmbeddingGemma 2: Unified and Efficient Multimodal Representations — https://arxiv.org/abs/2610.37364
  • [21▲] Channel Selection for Mixture-of-Experts Decoding Needs No Calibration Data — https://arxiv.org/abs/2610.37386