GPQA
Benchmarks
TaskDatasetSOTA resultTrendResults
GPQA-diamond OpenR1-Math Harder Subset
54Accuracy
3
GPQA-D
60.4% Avg@4
3
GPQA Diamond
0.405Average Accuracy
3
GPQA Diamond
69.19Exact Match
3
GPQA v1 (test)
34.3Accuracy
3
GPQA
80.6Score
2
GPQA Diamond
70Score
2
GPQA Diamond
99Quality
2
GPQA Diamond
0.02Cost (cents)
2
GPQA
5Positive Seeds Count
2
GPQA
33.33Accuracy
2
GPQA
37.2BA @20
2
GPQA
55.4Score
2
GPQA Diamond
22Labeling Budget
2
GPQA diamond
50Pass@1
2
GPQA Main
33.93Accuracy (GPQA Main)
2
GPQA
4SC Baseline AUC
2
GPQA Diamond
77.8SC K=4 Accuracy
2
GPQA Diamond
61.11Accuracy
2
GPQA
24.2pass@1
2
GPQA
198Total Score
2
GPQA
168.73TPS
2
GPQA Diamond
1.24Alpha Coefficient
2
GPQA 0-shot
29.4Log Accuracy (0-shot)
2
GPQA Diamond (198)
28.6Accuracy
2