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

Commonsense Question Answering on CSQA (Accuracy)

88.9Accuracy

Human

1.33224.06646.869.534Mar 5, 2026Mar 11, 2026Mar 18, 2026Mar 25, 2026Mar 31, 2026Apr 7, 2026Apr 14, 2026
Updated 3d ago

Evaluation Results

MethodLinks
2026.03
88.9
2026.03
84.54
2026.03
83.21
2026.03
82.1
2026.03
80.18
2026.03
80.01
2026.03
79.69
2026.03
78.5
2026.03
76.58
2026.03
76.42
2026.03
76.3
2026.03
76.2
2026.03
75.84
2026.03
75.7
2026.03
74.77
2026.03
74.53
2026.03
74.44
2026.03
74
2026.03
73.79
2026.03
73.38
2026.03
72.89
2026.03
71
2026.03
69.9
2026.03
69.3
2026.03
68.9
2026.03
68.2
2026.03
67.5
2026.03
67.3
2026.03
67
2026.03
66.3
2026.03
64.7
2026.03
64.2
2026.04
63.96
2026.03
63.9
2026.04
63.31
2026.03
63
2026.04
61.83
2026.03
61.7
2026.04
60.86
2026.03
59.6
2026.04
59.54
2026.04
54.3
2026.04
52.78
2026.04
51.25
2026.03
50
2026.03
48
2026.03
45
2026.03
44.2
2026.03
43
2026.03
41.4
2026.03
40.5
2026.03
32.4
2026.03
29.4
2026.03
25.4
2026.04
18.19
2026.04
12.1
2026.04
6.49
2026.04
4.7