GPQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
GPQA Diamond
72.22pass@1
12
GPQA
69,081Research And Data Score
12
GPQA (test)
83,717Research And Data Score
12
GPQA-D
59.59Pass@1
12
GPQA Diamond
100Quality (%)
11
GPQA
64.96Accuracy
11
GPQA Diamond
84.34Accuracy
11
GPQA
40.4Accuracy
11
GPQA Ambiguity-Augmented (subset of 200 samples)
44Accuracy@1
11
GPQA ambiguity-augmented
42.8Accuracy
11
GPQA Diamond
2,004.7Token Count
11
GPQA-D
100Coverage Rate
11
GPQA Main
20.8Accuracy
11
GPQA (test)
65.7Accuracy
11
GPQA
65.7Accuracy
11
GPQA
48.9Accuracy
11
GPQA
34.6Score
11
GPQA
42.41Solve Rate
11
GPQA Diamond
71.6AUCOAA
11
GPQA (test)
26.6AR
11
GPQA Diamond
69.7Score
11
GPQA-D
67.2Pass@1
11
GPQA diamond
87.37Average Score
11
GPQA n=546 (test)
100Act Score
10
GPQA
31.9Pass@1 Accuracy
10