General Competence Evaluation on MMLU-Pro
66.82AccuracySIA-15B
Evaluation Results
| Method | Links | |
|---|---|---|
| SIA-15B2026.03 | 66.82 | |
| Qwen2.5-14B-Instructvariant=baseline2026.03 | 63.7 | |
| 80/20Base Model=Qwen2.5-14B-Base2026.05 | 48.25 | |
| Ent-RegBase Model=Qwen2.5-14B-Base2026.05 | 48.15 | |
| DAPOBase Model=Qwen2.5-14B-Base2026.05 | 48 | |
| PAPOBase Model=Qwen2.5-14B-Base2026.05 | 47.89 | |
| BaseBase Model=Qwen2.5-14B-Base2026.05 | 47.33 | |
| DAPOBase Model=Qwen2.5-7B-Base2026.05 | 37.81 | |
| PAPOBase Model=Qwen2.5-7B-Base2026.05 | 37.81 | |
| Ent-RegBase Model=Qwen2.5-7B-Base2026.05 | 37.02 | |
| BaseBase Model=Qwen2.5-7B-Base2026.05 | 36.98 | |
| 80/20Base Model=Qwen2.5-7B-Base2026.05 | 36.92 |