List your company

Mathematical Reasoning via Verifiable Subquestions on IMProofBench

71.8Subquestion ScoreGemini-3-Pro
37.37646.31355.2564.187Feb 6, 2026

Evaluation results

MethodLinks
Gemini-3-Pro2026.0271.8
GPT-52026.0254.5
Claude-Opus-4.12026.0238.7