Automated Inequality Proving on PolyIneqBench n=8
21.7Pass RateNSPI
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| NSPI2026.05 | 21.7 | 63.9 | |
| Gemini-3-Pro2026.05 | 15 | 126.7 | |
| LIPS2026.05 | 8.3 | 153.1 | |
| Maple2026.05 | 6.67 | 456.3 | |
| DeepSeek-V3.22026.05 | 5 | 561.5 | |
| GPT-5.22026.05 | 3.3 | 55.9 | |
| Z32026.05 | 1.7 | 1,598.6 | |
| DS-Prover-v22026.05 | 0 | — | |
| Goedel-Prover-v22026.05 | 0 | — | |
| Kimina-Prover2026.05 | 0 | — |