Theorem Proving on VERINA basic
102Correct Solved Tests CountNumina Lean Agent
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Numina Lean AgentBenchmark variant=P2026.05 | 102 | 2 | |
| Claude Code + lean4-skillsBenchmark variant=P2026.05 | 96 | — | |
| AristotleBenchmark variant=PM2026.05 | 89 | 1 | |
| Kimina-ProverBenchmark variant=P2026.05 | 31 | — | |
| GrindBenchmark variant=PU2026.05 | 31 | — | |
| LeanHammerBenchmark variant=PU2026.05 | 13 | — | |
| AesopBenchmark variant=PU2026.05 | 13 | — | |
| CanonicalBenchmark variant=PU2026.05 | 0 | — |