Automated Inequality Proving on PolyIneqBench n=6
29.5Pass RateNSPI
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| NSPI2026.05 | 29.5 | 21.4 | |
| Gemini-3-Pro2026.05 | 21.3 | 138.9 | |
| Z32026.05 | 19.7 | 196.4 | |
| LIPS2026.05 | 11.5 | 195.7 | |
| Maple2026.05 | 8.2 | 54.9 | |
| GPT-5.22026.05 | 4.9 | 154.4 | |
| DeepSeek-V3.22026.05 | 1.6 | 495.8 | |
| DS-Prover-v22026.05 | 0 | — | |
| Goedel-Prover-v22026.05 | 0 | — | |
| Kimina-Prover2026.05 | 0 | — |