Reasoning on Mathematics (Normalized Score)
100Normalized ScoreMath expert
Evaluation results
| Method | Links | |
|---|---|---|
| Math expert2026.03 | 100 | |
| Oracle2026.03 | 94.3 | |
| DUME2026.03 | 91.1 | |
| DUME OOD2026.03 | 79.1 | |
| Model averaging2026.03 | 44.7 | |
| Random routing2026.03 | 39.4 | |
| Instr. following expert2026.03 | 28.2 | |
| Coding expert2026.03 | 12 | |
| Multilingual expert2026.03 | 6.1 |