《DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning》

(用的google翻译，凑和能看~~~，共22页，第17页起为贡献&致谢)

P1:

P2:

P3:

P4:

P5:

P6:

P7:

P8:

P9:

P10:

P11:

P12:

P13:

P14:

P15:

P16:

P17:

P18:

P19:

P20:

P21:

P22: