Skill Self-Play Makes Open-Ended Training Verifiable

Featured

• Skill Self-Play: Pushing the Frontier of LLM Capability with Co-Evolving Skills score 10
  入选 HF Daily Papers; HF 热度: 43 upvotes (+4); 有代码实现; 关键词(2): serving, reasoning
• Molt: A Scalable PyTorch-Native Training Framework for Agentic Reinforcement Learning score 10
  入选 HF Daily Papers; HF 热度: 29 upvotes (+4); 有代码实现; 关键词(3): agentic, coding, open source
• Scaling Native Multimodal Pre-Training From Scratch score 8
  入选 HF Daily Papers; HF 热度: 24 upvotes (+4); 关键词(4): scaling, pre-training, reasoning, vision-language
• IDEAgent: Agentic Quality-Diversity Search for Research Idea Generation score 8
  入选 HF Daily Papers; HF 热度: 8 upvotes (+2); 有代码实现; 关键词(3): lightweight, agentic, open-source
• SceneActBench: Can Agents Act on the 3D Scenes They See? score 8
  入选 HF Daily Papers; HF 热度: 6 upvotes (+2); 有代码实现; 关键词(1): vision-language
• Spectral Prior for Reducing Exposure Bias in Diffusion Models score 8
  入选 HF Daily Papers; HF 热度: 5 upvotes (+2); 有代码实现; 关键词(1): lightweight
• LAMAR: An Open Language-Aware Multilingual Alignment Reranker score 7
  入选 HF Daily Papers; HF 热度: 13 upvotes (+3); 关键词(1): distillation
• Closing the Loop: Training-Free Revisit Consistency for Autoregressive Generative Rendering score 6
  入选 HF Daily Papers; HF 热度: 6 upvotes (+2); 关键词(1): post-training