Mathematical Reasoning on AIME24 (P@4, Avg Output Length)
73.3P@4MPD
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| MPD2026.05 | 73.3 | 13.4 | |
| Direct Comp.2026.05 | 70 | 14.5 | |
| CRISP2026.05 | 70 | 14.8 | |
| LiteCoT2026.05 | 66.7 | 15.6 | |
| Vanilla LLM2026.05 | 63.3 | 16.8 | |
| Chain-of-Draft2026.05 | 63.3 | 11.8 |