马东锡 NLP 🇸🇪 0 关注者 关注 10个月前 「DeepSeek, Reasoning」论文 DeepSeek-Prover-V2: Advancing Formal Mathematical Reasoning via Reinforcement Learning for Subgoal Decomposition 用"sorry"做占位符，sorry，除了硬核，无法可说。 DeepSeek这篇在reasoning的追求上，到了一个让 #DeepSeek #reasoning #Formal Mathematical Reasoning #Reinforcement Learning #Subgoal Decomposition 前往原网页查看
