Question Answering on TruthfulQA (MC1, MC2)
38.9MC1 ScoreQwen2.5-7B
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Qwen2.5-7B2026.04 | 38.9 | 56.3 | |
| Qwen2.5-7B + MathAgentTraining Type=SFT, Data Size=1K2026.04 | 38.9 | 56 | |
| Qwen2.5-7B + NuminaMathTraining Type=SFT, Data Size=1K2026.04 | 38.8 | 55.9 | |
| Qwen2.5-7B + LIMOTraining Type=SFT, Data Size=0.8K2026.04 | 37.9 | 56.2 | |
| Qwen2.5-7B + MagpieTraining Type=SFT, Data Size=1K2026.04 | 37.7 | 54.9 | |
| Qwen2.5-7B + S1KTraining Type=SFT, Data Size=1K2026.04 | 37.2 | 55.1 |