MaxProof: Scaling Mathematical Proof with Generative-Verifier RL and Population-Level Test-Time Scaling
2026/06/11 by Jiacheng Chen, Xinyu Zhang, Shunkai Zhang +20 · 14 voices · 1 citation
#cs.LG #cs.AI #cs.CL
paper · pdf
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 engineered for low false-positive rate. These capabilities are merged into a single released M3 model. At test time, MaxProof treats the model as a generator, verifier, refiner, and ranker, searches over a population of candidate proofs, and returns one final proof through tournament selection. With MaxProof test-time scaling, the M3 model reaches 35/42 on IMO 2025 and 36/42 on USAMO 2026, exceeding the human gold-medal threshold on both.
Cited by
Discussions
- Maxproof [hn, 137 points, 13 comments]
- Maxproof https://arxiv.org/abs/2606.13473 (https://news.ycombinator.com/item?id=48503014) [bsky, 0 points, 0 comments]
- Maxproof [bsky, 0 points, 0 comments]
- Maxproof #HackerNews https://arxiv.org/abs/2606.13473 [bsky, 0 points, 0 comments]
- Maxproof https://arxiv.org/abs/2606.13473 https://news.ycombinator.com/item?id=48503014 [bsky, 0 points, 0 comments]
- Maxproof https://arxiv.org/abs/2606.13473 [bsky, 0 points, 0 comments]
- Maxproof https://arxiv.org/abs/2606.13473 [comments] [128 points] [bsky, 0 points, 0 comments]
- 📰 Maxproof 🔗 https://arxiv.org/abs/2606.13473 💬 Discuss on HN [bsky, 0 points, 0 comments]
- Maxproof https://arxiv.org/abs/2606.13473 (https://news.ycombinator.com/item?id=48503014) [bsky, 0 points, 0 comments]
- Links: abs, pdf Search: Bluesky, Twitter, Reddit, Hacker News, Hugging Face, alphaXiv [bsky, 0 points, 1 comments]
- 4/ 📄 MaxProof. Generative-verifier RL plus population-level test-time scaling, reportedly past human gold-medal thresholds on IMO 2025 and USAMO 2026. Worth reading honestly for the test-time scaling [bsky, 0 points, 0 comments]
- 📰 Maxproof, an AI system designed to detect and prevent malicious software from entering systems, has been developed by researchers at MIT's Computer Science and Artificial Intelligence Laboratory (C [bsky, 0 points, 0 comments]
- Maxproof https:// arxiv.org/abs/2606.13473 # arxiv [mastodon, 0 points, 0 comments]
- Maxproof https:// arxiv.org/abs/2606.13473 # HackerNews # Maxproof # research # arxiv # tech # innovation # security [mastodon, 0 points, 0 comments]
Related