结构推理拿67项SOTA，异步RL入GLM-5.2

重点关注

• Accurate, Interdisciplinary and Transparent Structure-property Understanding with Deep Native Structural Reasoning score 10
  入选 HF Daily Papers；HF 热度: 83 upvotes (+4)；有代码实现；关键词(1): reasoning
• AgentLens: Production-Assessed Trajectory Reviews for Coding Agent Evaluation score 8
  入选 HF Daily Papers；HF 热度: 7 upvotes (+2)；有代码实现；关键词(3): production, coding, open source
• WildCity: A Real-World City-Scale Testbed for Rendering, Simulation, and Spatial Intelligence score 8
  入选 HF Daily Papers；HF 热度: 8 upvotes (+2)；有代码实现；关键词(2): scaling, embodied
• Single-Rollout Asynchronous Optimization for Agentic Reinforcement Learning score 7
  入选 HF Daily Papers；HF 热度: 13 upvotes (+3)；关键词(6): throughput, GRPO, post-training, agentic, coding
• Sparse Delta Memory: Scaling the State of Linear RNNs through Sparsity score 7
  入选 HF Daily Papers；HF 热度: 10 upvotes (+3)；关键词(2): scaling, reasoning

也值得关注

• UP: Unbounded Positive Asymmetric Optimization for Breaking the Exploration-Stability Dilemma score 5
  入选 HF Daily Papers；HF 热度: 4 upvotes (+1)；关键词(4): GRPO, MoE, reasoning, vision-language