Computer Science > Machine Learning Title:MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling View PDF HTML (experimental)Abstract:We present MaxProof, a population-level test-time scaling framework for competition-level mathematical proof in the MiniMax-M3 series. M3 first trains three proof-oriented capabilities -- proof generation, proof verification, and critique-conditioned proof repair -- using a defense-in-depth generative verifier...

Maxproof
Jiacheng; Zhang; Xinyu; Shunkai; Wang; Yanmohan; Lin; Qin; Tiancheng; Zhu; Zhengmao; Tianle; Jingyang; Zehan; Jiang; Binyang; Ding; Han; Fei; Du; Chenyu; Song; Zijian; Jiayuan; Zhi; Huang; Yunan; Cheng; Weiyu; Zhao; Pengyu
