GPQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
GPQA
52Pass@1
8
GPQA
67.8AUROC
8
GPQA
43.43Average Score (@8)
8
GPQA
33.3Avg@16 Score
8
GPQA
51.01Overall Score
8
GPQA Diamond
27.32Relative MAE
8
GPQA Diamond
13.22rMAE
8
GPQA
15.04Throughput
8
GPQA-D (hold-out)
76.82Accuracy
8
GPQA-D
66.5Accuracy
8
GPQA
90Research & Data Score
8
GPQA
81.25Task Completion Rate
8
GPQA
65Task Completion Rate
8
GPQA
75.86Task Completion Rate
8
GPQA Diamond
90.7Avg@16
8
GPQA Think (test)
61.1Accuracy
8
GPQA No-think (test)
47.8Accuracy
8
GPQA 0-shot (test)
28.57Pass@1 Accuracy
8
GPQA D
67.4Accuracy (Avg@8)
8
GPQA
36.33RA
8
GPQA
32.8Genetics Accuracy
7
GPQA
51.8Accuracy
7
GPQA 5-shot
34.75Accuracy
7
GPQA
32.25Accuracy
7
GPQA Diamond
61.6Accuracy
7