Chain-of-Thought Reasoning on CoT-Collection
73.7Composite ScoreSFT-Tag
Evaluation Results
| Method | Links | |
|---|---|---|
| SFT-TagModel=Qwen2.5-7B2025.12 | 73.7 | |
| SFT(baseline)Model=Qwen2.5-7B2025.12 | 73.48 | |
| SFTKey-TagModel=Qwen2.5-7B2025.12 | 73.42 | |
| SFTKey-TagModel=Qwen3-8B-Base2025.12 | 72.98 | |
| SFT(baseline)Model=Qwen3-8B-Base2025.12 | 72.78 | |
| SFT-TagModel=Qwen3-8B-Base2025.12 | 72.64 | |
| Key-TagModel=Qwen3-8B-Base2025.12 | 69.21 | |
| SFTKey-TagModel=SmolLM3-3B-Base2025.12 | 67.97 | |
| SFT(baseline)Model=SmolLM3-3B-Base2025.12 | 66.78 | |
| SFT-TagModel=SmolLM3-3B-Base2025.12 | 65.02 | |
| Key-TagModel=Qwen2.5-7B2025.12 | 44.8 | |
| SFTKey-TagModel=Qwen2.5-3B2025.12 | 39.97 | |
| SFT-TagModel=Qwen2.5-3B2025.12 | 37.52 | |
| SFT(baseline)Model=Qwen2.5-3B2025.12 | 37.24 | |
| Key-TagModel=SmolLM3-3B-Base2025.12 | 36.94 | |
| Key-TagModel=Qwen2.5-3B2025.12 | 34.3 | |
| Key-TagModel=Qwen2.5-1.5B2025.12 | 34.3 | |
| SFT-TagModel=Qwen2.5-1.5B2025.12 | 33.32 | |
| SFT(baseline)Model=Qwen2.5-1.5B2025.12 | 32.48 | |
| SFTKey-TagModel=Qwen2.5-1.5B2025.12 | 31.92 |