Mathematical Reasoning on MATH500 (Accuracy, Pearson r, R2)
0.691Pearson rICRM
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ICRMRegularization (lambda)=0.12026.02 | 0.691 | 0.477 | 0.459 | 45.4 | |
| ICRMRegularization (lambda)=0.52026.02 | 0.685 | 0.469 | 0.461 | — | |
| Bradley-Terry2026.02 | 0.663 | 0.439 | 0.428 | 44.6 | |
| Exact Match2026.02 | — | — | — | 45 |