Scientific Question Answering on GPQA (Domain Scores)
65.8Biology Domain ScoreSkywork
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| SkyworkBase Model=Qwen3-8B-Base, Metric=Mean@N2026.05 | 65.8 | 52.9 | 41.9 | 49 | |
| DAPO++Base Model=Qwen3-8B-Base, Metric=Mean@N2026.05 | 65.8 | 68 | 41.7 | 55.4 | |
| OpenR1Base Model=Qwen3-8B-Base, Metric=Mean@N2026.05 | 61.8 | 39.8 | 40.9 | 42.4 | |
| Qwen3-8B-BaseModel Size=8B, Metric=Mean@N2026.05 | 60.5 | 45.9 | 37.9 | 43.6 | |
| DeepMathBase Model=Qwen3-8B-Base, Metric=Mean@N2026.05 | 60.5 | 58.7 | 33.3 | 47 | |
| DeepScaleRBase Model=Qwen3-8B-Base, Metric=Mean@N2026.05 | 56.6 | 47.1 | 33.9 | 41.8 | |
| DAPOBase Model=Qwen3-8B-Base, Metric=Mean@N2026.05 | 51.3 | 68.9 | 40.9 | 54 | |
| DeepMathBase Model=Qwen3-1.7B-Base, Metric=Mean@N2026.05 | 48.7 | 39.2 | 42.5 | 41.7 | |
| DAPOBase Model=Qwen3-1.7B-Base, Metric=Mean@N2026.05 | 44.7 | 33.1 | 35.8 | 35.5 | |
| DeepScaleRBase Model=Qwen3-1.7B-Base, Metric=Mean@N2026.05 | 43.4 | 29.7 | 34.1 | 33.1 | |
| DAPO++Base Model=Qwen3-1.7B-Base, Metric=Mean@N2026.05 | 42.1 | 34 | 36.3 | 35.9 | |
| SkyworkBase Model=Qwen3-1.7B-Base, Metric=Mean@N2026.05 | 36.8 | 35.2 | 38.2 | 36.7 | |
| OpenR1Base Model=Qwen3-1.7B-Base, Metric=Mean@N2026.05 | 34.2 | 27.3 | 33.9 | 31.1 | |
| Qwen3-1.7B-BaseModel Size=1.7B, Metric=Mean@N2026.05 | 32.9 | 28.5 | 30.1 | 29.7 |