思而不学则殆：通义实验室×北大联合提出RL-PLUS，突破大模型推理边界
"复杂推理" 相关文章
TPO: Aligning Large Language Models with Multi-branch & Multi-step Preference Trees
智能体系统如何「边做边学」？斯坦福团队探索在线优化的新范式
智能体系统如何「边做边学」？斯坦福团队探索在线优化的新范式
软件所提出基于信息论的大模型强化学习微调框架
Scaf-GRPO: Scaffolded Group Relative Policy Optimization for Enhancing LLM Reasoning
SmartSwitch: Advancing LLM Reasoning by Overcoming Underthinking via Promoting Deeper Thought Exploration
Gemini 3.0 匿名上线
Budget-aware Test-time Scaling via Discriminative Verification
IMAGINE: Integrating Multi-Agent System into One Model for Complex Reasoning and Planning
MULTI: Multimodal Understanding Leaderboard with Text and Images
What Makes Looped Transformers Perform Better Than Non-Recursive Ones (Provably)
What Makes Looped Transformers Perform Better Than Non-Recursive Ones (Provably)
蚂蚁开源万亿参数大模型Ling-1T：多项能力全球领先
Samsung’s tiny AI model beats giant reasoning LLMs
Samsung’s tiny AI model beats giant reasoning LLMs
Samsung’s tiny AI model beats giant reasoning LLMs
What Can You Do When You Have Zero Rewards During RL?
What Can You Do When You Have Zero Rewards During RL?
Can an LLM Induce a Graph? Investigating Memory Drift and Context Length
