The Model as Its Own Teacher, and Why Splitting Crowds Wins

Featured

• SEED: Self-Evolving On-Policy Distillation for Agentic Reinforcement Learning score 10
  入选 HF Daily Papers; HF 热度: 78 upvotes (+4); 有代码实现; 关键词(3): distillation, agentic, tool use
• SearchOS-V1: Towards Robust Open-Domain Information-Seeking Agent Collaboration score 10
  入选 HF Daily Papers; HF 热度: 54 upvotes (+4); 有代码实现; 关键词(1): throughput
• RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination score 9
  入选 HF Daily Papers; HF 热度: 19 upvotes (+3); 有代码实现; 关键词(4): pretraining, reasoning, vision-language, embodied
• Partition, Prompt, Aggregate: Statistical Self-Consistency in Language Models score 6
  入选 HF Daily Papers; HF 热度: 4 upvotes (+1); 有代码实现
• Chat2Scenic: An Iterative RAG-Based Framework for Scenario Generation in Autonomous Driving score 6
  入选 HF Daily Papers; 有代码实现; 关键词(3): retrieval-augmented, RAG, open source
• Smarter and Cheaper at Once: Byte-Exact KV-Cache Grafting Turns a Frozen Small Model into a Verified-Knowledge Flywheel score 5
  入选 HF Daily Papers; HF 热度: 6 upvotes (+2)

Also Worth Noting

• Token Time Continuous Diffusion for Language Modeling score 4
  入选 HF Daily Papers; HF 热度: 2 upvotes (+1)