Autoformalization and Proving on ProofNet N=186 (test)
0.7849Pass@64Ground truth statement (oracle)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Ground truth statement (oracle)2026.03 | 0.7849 | 0.2473 | 0.1828 | |
| FormalEvolveK=22026.03 | 0.6828 | 0.2796 | 0.2419 | |
| Compile+Semantic RepairSource=Kimina2026.03 | 0.6398 | 0.2688 | 0.2473 | |
| Sample2026.03 | 0.5699 | 0.2473 | 0.2204 |