T²PO稳多轮RL+视频缓存提速6倍

重点关注

• AcademiClaw: When Students Set Challenges for AI Agents score 8
  入选 HF Daily Papers；HF 热度: 10 upvotes (+3)；有代码实现
• Hallucinations Undermine Trust; Metacognition is a Way Forward score 7
  入选 HF Daily Papers；HF 热度: 18 upvotes (+3)；关键词(1): agentic
• PhysicianBench: Evaluating LLM Agents in Real-World EHR Environments score 8
  入选 HF Daily Papers；HF 热度: 8 upvotes (+2)；有代码实现；关键词(2): reasoning, open-source
• T$^2$PO: Uncertainty-Guided Exploration Control for Stable Multi-Turn Agentic Reinforcement Learning score 11
  入选 HF Daily Papers；HF 热度: 6 upvotes (+2)；有代码实现；关键词(2): agentic, reasoning；顶会接收: ICML
• Counting as a minimal probe of language model reliability score 7
  入选 HF Daily Papers；HF 热度: 3 upvotes (+1)；有代码实现；关键词(2): coding, reasoning
• Motion-Aware Caching for Efficient Autoregressive Video Generation score 8
  入选 HF Daily Papers；HF 热度: 5 upvotes (+2)；有代码实现；关键词(2): lightweight, deployment
• Linear-Time Global Visual Modeling without Explicit Attention score 6
  入选 HF Daily Papers；HF 热度: 4 upvotes (+1)；有代码实现
• Linking spatial biology and clinical histology via Haiku score 5
  入选 HF Daily Papers；有代码实现
• A Hybrid Approach for Closing the Sim2real Appearance Gap in Game Engine Synthetic Datasets score 5
  入选 HF Daily Papers；有代码实现

也值得关注

• Agentic AI Systems Should Be Designed as Marginal Token Allocators score 5
  入选 HF Daily Papers；HF 热度: 3 upvotes (+1)；关键词(4): serving, latency, agentic, coding
• Uncertainty-Aware Trip Purpose Inference from GPS Trajectories via POI Semantic Zones and Pareto Calibration score 3
  机构: Stanford
• GeoSAE: Geometric Prior-Guided Layer-Wise Sparse Autoencoder Annotation of Brain MRI Foundation Models score 3
  机构: Stanford