Graduate-level Question Answering on GPQA Diamond (test)
37.4AccuracyAnchor-Points
Evaluation Results
| Method | Links | |
|---|---|---|
| Anchor-PointsOptimization Budget=Large2026.01 | 37.4 | |
| SESS-wrepOptimization Budget=Small2026.01 | 37.4 | |
| SESS-vlcOptimization Budget=Large2026.01 | 36.9 | |
| RandomOptimization Budget=Large2026.01 | 35.9 | |
| SESS-wrepOptimization Budget=Large2026.01 | 34.8 | |
| RandomOptimization Budget=Small2026.01 | 34.3 | |
| SESS-lcOptimization Budget=Small2026.01 | 34.3 | |
| SESS-vlcOptimization Budget=Small2026.01 | 34.3 | |
| IPOMPOptimization Budget=Small2026.01 | 33.8 | |
| IPOMPOptimization Budget=Large2026.01 | 33.8 | |
| Anchor-PointsOptimization Budget=Small2026.01 | 33.8 | |
| SESS-lcOptimization Budget=Large2026.01 | 32.8 | |
| SESS-repOptimization Budget=Small2026.01 | 32.3 | |
| BaseOptimization Budget=Small2026.01 | 29.9 | |
| BaseOptimization Budget=Large2026.01 | 29.9 | |
| SESS-repOptimization Budget=Large2026.01 | 29.8 |