Mathematical Reasoning on AIME Combined (Multilingual Scores)
72.3English ScoreQwen3-8B (Thinking)
Evaluation Results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| Qwen3-8B (Thinking)Base Model=Qwen3-8B, Target-CoT setting=true, Thinking mode=true2025.08 | 72.3 | 54.3 | 49.3 | 32.7 | 38 | 37.3 | 26 | 4.3 | 4.3 | 1 | 27.5 | |
| OpenThoughts3-20kSFT Dataset=OpenThoughts3-20k, Base Model=Qwen3-8B-Base, Target-CoT setting=true, Thinking mode=false2025.08 | 42.3 | 27 | 31 | 21.3 | 23.3 | 22.3 | 12.7 | 5 | 3.3 | 0.3 | 16.2 | |
| Translated-s1kSFT Dataset=Translated-s1k, Base Model=Qwen3-8B-Base, Target-CoT setting=true, Thinking mode=false2025.08 | 33 | 29.3 | 29 | 16 | 27.7 | 19.3 | 24.7 | 14.7 | 15.3 | 11.3 | 20.8 | |
| Distilled-s1kSFT Dataset=Distilled-s1k, Base Model=Qwen3-8B-Base, Target-CoT setting=true, Thinking mode=false2025.08 | 33 | 10.7 | 21 | 13 | 28.3 | 19.7 | 22 | 16.3 | 13.7 | 12.7 | 17.5 |