GPQA
Benchmarks
Task NameDataset NameSOTA ResultTrendResults
GPQA Diamond (test)
34.85Accuracy
4
GPQA
54.7Accuracy
4
GPQA-D
23.4Accuracy
4
GPQA D
27.3Accuracy
4
GPQA
25.8Accuracy
4
GPQA
27Accuracy
4
GPQA Diamond
68.18Pass@3
4
GPQA Diamond
84.2Pass@1 Score
4
GPQA
33Log Accuracy
4
GPQA Diamond
39.87#Allocs
4
GPQA Diamond In-Distribution (test)
80.36CPT (90%)
4
GPQA-Diamond OOD
74.3CPT Accuracy (85%)
4
GPQA Diamond
89.19CPT (95%)
4
GPQA Diamond
82.54CPT (90%)
4
GPQA Diamond (ID)
62.69CPT (80%)
4
GPQA Diamond (ID queries)
71.53CPT (85%)
4
GPQA Diamond
90.78CPT (95%)
4
GPQA Diamond
0.747Hypervolume
4
GPQA Diamond OOD
66.05CPT (80%)
4
GPQA-Diamond (test)
0.751Hypervolume
4
GPQA Diamond
31.5Avg@8
4
GPQA
31.7Inference Time (s)
4
GPQA
27Score
4
GPQA
66.67Accuracy
4
GPQA
19.7ANLL
4