Jasper Dekoninck(@j_dekoninck):We just added a new entry to ArXivLean! The Mistral team ran an agent consisting of two open models (Leanstral 1.5 as prover, Kimi K3 as coordinator) and obtained a massive 82% on the leaderboard, using an equally massive 10 million output tokens per problem.

Jasper Dekoninck

@j_dekoninck

PhD Student at ETH Zurich, working on mathematical evaluation of LLMs (MathArena, IMproofBench, BrokenMath, ...)

加入 June 2023

136 正在關注 4.4K 粉絲

Jasper Dekoninck@j_dekoninck

2026.09.11 12:14

We just added a new entry to ArXivLean! The Mistral team ran an agent consisting of two open models (Leanstral 1.5 as prover, Kimi K3 as coordinator) and obtained a massive 82% on the leaderboard, using an equally massive 10 million output tokens per problem.