Mathematical Reasoning on AIME24 (Pass@1, #Tk)
33.33Pass@1DRP
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| DRPBackbone=Phi-4-Reasoning2026.01 | 33.33 | 6,135 | |
| TALEBackbone=Phi-4-Reasoning2026.01 | 29 | 2,069.97 | |
| FROSTBackbone=Phi-4-Reasoning2026.01 | 26.67 | 899.8 | |
| DRPBackbone=GPT-oss-20B2026.01 | 22.45 | 4,983 | |
| TALEBackbone=GPT-oss-20B2026.01 | 20 | 1,354.67 | |
| FROSTBackbone=GPT-oss-20B2026.01 | 16.67 | 1,009.6 | |
| ThinkLessBackbone=Phi-4-Reasoning2026.01 | 16.08 | 1,405.4 | |
| ThinkLessBackbone=GPT-oss-20B2026.01 | 16 | 1,205 | |
| SelfBudgeterBackbone=Phi-4-Reasoning2026.01 | 13.42 | 1,372.83 | |
| BaseBackbone=GPT-oss-20B2026.01 | 13.33 | 1,003.57 | |
| SelfBudgeterBackbone=GPT-oss-20B2026.01 | 13.2 | 1,256 | |
| FROSTBackbone=Magistral-Small-1.12026.01 | 9.74 | 149.93 | |
| BaseBackbone=Phi-4-Reasoning2026.01 | 6.67 | 1,017.7 | |
| ThinkLessBackbone=Magistral-Small-1.12026.01 | 6 | 1,205 | |
| SelfBudgeterBackbone=Magistral-Small-1.12026.01 | 5.2 | 1,256 | |
| DRPBackbone=Magistral-Small-1.12026.01 | 4.5 | 1,350.77 | |
| TALEBackbone=Magistral-Small-1.12026.01 | 3.33 | 967.43 | |
| BaseBackbone=Magistral-Small-1.12026.01 | 0 | 537.13 |