Mathematical Reasoning on AMC 23 (Acc, Tok, EPR, Avg. EPR)
90.3AccuracyVanilla
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| VanillaModel=DS-7B2026.01 | 90.3 | 6,434 | — | — | |
| EntrocutModel=DS-7B2026.01 | 89.2 | 5,216 | 3.4 | 2.4 | |
| TIPModel=DS-7B2026.01 | 87.3 | 4,714 | 1.8 | 1.3 | |
| DEERModel=DS-7B2026.01 | 86.9 | 4,322 | 1.9 | 1.5 | |
| NOWAITModel=DS-7B2026.01 | 82.2 | 3,811 | — | — | |
| VanillaModel=DS-1.5B2026.01 | 73 | 8,834 | — | — | |
| EntrocutModel=DS-1.5B2026.01 | 72.8 | 6,000 | 25.7 | 12.4 | |
| TIPModel=DS-1.5B2026.01 | 72 | 5,998 | 5.1 | 2.9 | |
| DEERModel=DS-1.5B2026.01 | 68.2 | 5,152 | 1.4 | 1.7 | |
| NOWAITModel=DS-1.5B2026.01 | 65.3 | 4,594 | — | — |