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

Scientific Question Answering on GPQA (accuracy)

54.44Accuracy

Spurious

22.470430.770239.0747.3698May 28, 2026Jun 3, 2026Jun 10, 2026Jun 17, 2026Jun 23, 2026Jun 30, 2026Jul 7, 2026
Updated 17d ago

Evaluation Results

MethodLinks
2026.05
54.44
2026.05
50.8
2026.05
49.59
2026.07
49.24
2026.07
44.44
2026.07
44.3
2026.07
43.94
2026.07
42.87
2026.05
41.11
2026.07
40.4
2026.07
40.2
2026.05
40
2026.05
37.78
2026.05
37.78
2026.07
37.12
2026.05
36.67
2026.05
36.67
2026.05
35.56
2026.05
35.56
2026.05
35.56
2026.05
34.44
2026.05
34.44
2026.05
33.33
2026.05
33.33
2026.05
33.33
2026.05
33.33
2026.05
33.33
2026.05
33.33
2026.05
33.33
2026.07
33.3
2026.07
33
2026.05
32.22
2026.05
31.85
2026.05
31.11
2026.07
30.8
2026.05
30
2026.05
30
2026.05
28.93
2026.05
28.89
2026.05
27.78
2026.05
27.78
2026.05
27.43
2026.05
26.67
2026.05
26.67
2026.05
26.3
2026.05
25.56
2026.05
25.56
2026.05
25.56
2026.05
24.44
2026.05
24.07
2026.05
23.7