Mathematical Reasoning on AIME Aya (Reduced)
63.2AccuracyModel-first Greedy
Evaluation Results
| Method | Links | |
|---|---|---|
| Model-first Greedyk=5, Summarizer=Aya2026.05 | 63.2 | |
| Input-allk=5, Summarizer=Aya2026.05 | 62.1 | |
| MoAk=5, Summarizer=Aya2026.05 | 56.2 | |
| Truth-prediction Greedyk=5, Summarizer=Aya2026.05 | 52.2 | |
| Oracle-surrogate Greedyk=5, Summarizer=Aya2026.05 | 49.7 | |
| Conditioned-diversityk=5, Summarizer=Aya2026.05 | 46.8 | |
| GPT5.2-judgek=5, Summarizer=Aya2026.05 | 45 | |
| Aya-judgek=5, Summarizer=Aya2026.05 | 41.5 | |
| Top-accuracyk=5, Summarizer=Aya2026.05 | 36.4 | |
| Best-modelk=5, Summarizer=Aya2026.05 | 33.2 |