Scientific Reasoning on GPQA Diamond (Subject Scores)
58.6Biology ScoreDeepMath-Zero-7B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| DeepMath-Zero-7BParameters=7B2025.10 | 58.6 | 29.5 | 53.2 | 42.6 | |
| DeepCompress-Zero-7BParameters=7B2025.10 | 57.6 | 31.2 | 58.2 | 43.9 | |
| Open-Reasoner-Zero-7BParameters=7B2025.10 | 50.3 | 26.7 | 47.8 | 38.1 | |
| Qwen-2.5-3B-InstructParameters=3B, Variant=Instruct2025.10 | 45.1 | 26.1 | 30.7 | 29.9 | |
| DeepMath-Zero-3BParameters=3B2025.10 | 45.1 | 25.3 | 32.2 | 30.2 | |
| DeepCompress-Zero-3BParameters=3B2025.10 | 44.1 | 25.5 | 35.7 | 31.7 | |
| Qwen-2.5-7BParameters=7B2025.10 | 33.6 | 21.4 | 27.8 | 25.3 | |
| Qwen-2.5-7B-SimpleRL-ZooParameters=7B, Variant=SimpleRL-Zoo2025.10 | 31.9 | 22.6 | 37.9 | 30.2 | |
| Qwen-2.5-3BParameters=3B2025.10 | 29.9 | 19.8 | 20.3 | 21 |