Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Multiple Choice Questions on GPQA

93.5Accuracy

BERT-Judge

24.65242.52660.478.274Mar 22, 2026Mar 25, 2026Mar 28, 2026Mar 31, 2026Apr 3, 2026Apr 6, 2026Apr 10, 2026
Updated 5d ago

Evaluation Results

MethodLinks
2026.04
93.5--
2026.04
86.5--
2026.04
66.2--
2026.03
54.175.161.2
2026.03
53.274.359.9
2026.03
52.172.860.1
2026.03
49.773.158.5
2026.03
4872.255.2
2026.03
46.968.954.5
2026.03
42.372.456.7
2026.03
40.463.243.8
2026.03
40.459.946.7
2026.03
34.654.334.2
2026.03
34.456.938.5
2026.03
33.549.832.9
2026.03
32.768.746.6
2026.03
32.548.230.4
2026.03
32.356.131.8
2026.03
32.157.233.2
2026.03
3149.127.4
2026.03
29.555.629.7
2026.03
29.149.729.2
2026.03
28.348.226.5
2026.03
27.34027.2