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

Science Question Answering on SciQ (Accuracy)

85.1Accuracy (SciQ)

Oracle

52.3460.84569.3577.855Mar 24, 2026Mar 27, 2026Mar 31, 2026Apr 3, 2026Apr 7, 2026Apr 10, 2026Apr 14, 2026
Updated 3d ago

Evaluation Results

MethodLinks
2026.04
85.1
2026.04
84.6
2026.04
82
2026.04
80.8
2026.04
79.8
2026.04
74.2
2026.04
74.2
2026.04
74.1
2026.04
74
2026.04
73.9
2026.04
73.9
2026.04
73.6
2026.04
72.9
2026.04
70.5
2026.04
70.4
2026.04
70.3
2026.04
70.3
2026.04
70.2
2026.04
70.1
2026.04
70
2026.04
69
2026.04
68.1
2026.04
68
2026.04
67.9
2026.03
67.8
2026.04
67.6
2026.04
67.3
2026.03
66.8
2026.04
66.8
2026.04
66.4
2026.04
66.3
2026.04
66.2
2026.04
65.6
2026.04
65.5
2026.04
65.2
2026.04
65.2
2026.04
64.8
2026.04
64.4
2026.04
64
2026.04
61.5
2026.04
61.1
2026.04
61
2026.04
60
2026.04
59.8
2026.04
59.6
2026.04
59.5
2026.04
59.4
2026.04
59.3
2026.04
59.2
2026.04
58.8
2026.04
53.6
2026.04
53.6