List your company

Theorem Proving on PutnamBench (test)

72AccuracyHilbert
-2.8816.563655.44Oct 14, 2025

Evaluation results

MethodLinks
HilbertCompute=avg. pass@1840, Open-source=false2025.1072462
Seed-ProverCompute=medium, Open-source=false2025.1051329
Ax-ProverCompute=pass@1‡, Open-source=true2025.101492
Goedel-Prover-V2Compute=pass@184, Open-source=true2025.101386
DeepSeek-Prover-V2Compute=pass@1024, Open-source=true2025.10747
DSP+Compute=pass@128, Open-source=true2025.10423
BourbakiCompute=pass@512, Open-source=true2025.10214
Kimina-Prover-7B-DistillCompute=pass@192, Open-source=true2025.10210
Self-play Theorem ProverCompute=pass@3200, Open-source=true2025.1018
Goedel-Prover-SFTCompute=pass@512, Open-source=true2025.1017
Gemini-2.5-ProCompute=pass@1, Open-source=false2025.100.53
GPT-4oCompute=pass@10, Open-source=false2025.100.21
Claude-3.7-SonnetCompute=pass@1, Open-source=false2025.1000