Multilingual Mathematical Reasoning on MCLM 1.0 (test)
84.89MT-MATH100 Scoreo3-Mini
Evaluation Results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| o3-Mini2025.02 | 84.89 | 45.33 | 29.75 | 51.42 | 52.85 | |
| GPT-4o-Mini2025.02 | 70.3 | 20.18 | 13.33 | 30.81 | 33.66 | |
| Deepseek-R1-1.5B (MR1-1.5B)Training Configuration=MT-SFT2025.02 | 55.61 | 19.94 | 19.2 | 28.97 | 30.93 | |
| Deepseek-R1-1.5B2025.02 | 49.4 | 17.21 | 21.94 | 26.77 | 28.83 | |
| Qwen2.5-Math-1.5B-Instruct2025.02 | 42.32 | 16.36 | 12.23 | 25 | 23.98 | |
| Qwen2.5-Math-1.5BTraining Configuration=MT-SFT2025.02 | 42.02 | 16.67 | 10.52 | 19.92 | 22.28 | |
| Qwen2.5-Math-1.5BTraining Configuration=SFT2025.02 | 37.47 | 14.85 | 10.5 | 18.4 | 20.3 |