Mathematical Reasoning on MATH (Pass@1, Pass@2)
88.42Pass@1Phi-4-mini + Mistral3-3B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Phi-4-mini + Mistral3-3BTraining Framework=CORE, Model Configuration=Team2026.01 | 88.42 | 92.08 | |
| Phi-4-mini-reasoningTraining Framework=CORE, Model Configuration=Single2026.01 | 85.24 | 90.01 | |
| Mistral3-3B-ReasoningTraining Framework=CORE, Model Configuration=Single2026.01 | 83.72 | 89.12 | |
| Phi-4-mini + Mistral3-3B + OracleTraining Framework=SD-E2, Model Configuration=Oracle2026.01 | 71.85 | 77.7 | |
| Phi-4-mini-reasoningTraining Framework=SD-E2, Model Configuration=Single2026.01 | 69.68 | 73.82 | |
| Mistral3-3B-ReasoningTraining Framework=SD-E2, Model Configuration=Single2026.01 | 67.25 | 73.15 | |
| Phi-4-mini + Mistral3-3B + OracleTraining Framework=Base, Model Configuration=Oracle2026.01 | 59.25 | 70.7 | |
| Phi-4-mini-reasoningTraining Framework=Base, Model Configuration=Single2026.01 | 57.04 | 68.16 | |
| Mistral3-3B-ReasoningTraining Framework=Base, Model Configuration=Single2026.01 | 54.21 | 64.44 |