HF Daily Papers · 2026-11-03

HF Daily Papers 精选（15 篇，按社区票数排序）

  • [112▲] FrameForge: Harnessing Priors of Video Foundation Models for Unified Image Generation — https://arxiv.org/abs/2610.36545
  • [89▲] R^2Orch: Recursive Self-Improvement for Agent Orchestration via Self-Play within a Skill-Internalization Loop — https://arxiv.org/abs/2610.22740
  • [71▲] Proving and Improving the Platonic Representation Hypothesis with JEPAs — https://arxiv.org/abs/2610.39302
  • [59▲] Does Scaling Toward AGI Depend on What System-1 Lets System-2 Execute? — https://arxiv.org/abs/2610.39497
  • [44▲] REA: A Routed Embodied Agent Runtime for Asynchronous Brain-Cerebellum Control — https://arxiv.org/abs/2610.42053
  • [34▲] Apriel Voice 27B Preview: An Open Model for Speech, Reasoning, and Tool Use — https://arxiv.org/abs/2610.40738
  • [26▲] WeFair: Fidelity-Aware Alignment from Human Feedback for Personalized Query Rewriting in Short-Video Search — https://arxiv.org/abs/2610.16927
  • [25▲] Follow the Gradient: Tracing Safety Degradation Back to Training Data — https://arxiv.org/abs/2610.37481
  • [25▲] Latent Reasoning in Continuous Space for Unified Multimodal Models — https://arxiv.org/abs/2610.42149
  • [24▲] Wizard: Manipulate Any World in Real Time via Agentic Control — https://arxiv.org/abs/2610.40339
  • [24▲] HiFOR: Hierarchical Foresight with Action-Guided World Grounding for Vision-Language-Action Model — https://arxiv.org/abs/2610.40373
  • [23▲] OSWAM: An Effective World Action Model with One-Step Action Generation — https://arxiv.org/abs/2610.35850
  • [22▲] ReLift: Dense 3D Tracking as Lifted Trajectory Repair — https://arxiv.org/abs/2610.42136
  • [20▲] Venus-DefakerOne2: A Vision-Forensics Foundation Model with Broader Coverage — https://arxiv.org/abs/2610.37918
  • [20▲] CEPO: A General Framework for Improving Prompt Optimization in Multi-Agent Systems — https://arxiv.org/abs/2610.34722