General Reasoning on GPQA (GPQA Metric)
35.04GPQA AccuracyINSIGHT
Evaluation Results
| Method | Links | |
|---|---|---|
| INSIGHTModel=Qwen3-4B2026.03 | 35.04 | |
| RANDOMModel=Qwen3-4B2026.03 | 34.38 | |
| EXPECTED-DIFFICULTYModel=Qwen3-4B2026.03 | 33.5 | |
| INVERSE-EVIDENCEModel=Qwen3-4B2026.03 | 33.48 | |
| MOPPSModel=Qwen3-4B2026.03 | 33.26 | |
| INSIGHTModel=Qwen3-1.7B2026.03 | 30.36 | |
| RANDOMModel=Qwen3-1.7B2026.03 | 30.2 | |
| MOPPSModel=Qwen3-1.7B2026.03 | 30.1 | |
| INVERSE-EVIDENCEModel=Qwen3-1.7B2026.03 | 29.69 | |
| INSIGHTModel=Qwen3-0.6B2026.03 | 29.5 | |
| INVERSE-EVIDENCEModel=Qwen3-0.6B2026.03 | 29.46 | |
| MOPPSModel=Qwen3-0.6B2026.03 | 27.9 | |
| RANDOMModel=Qwen3-0.6B2026.03 | 26.34 |