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

Hallucination Detection on CSQA

85.1AUROC

OSCAR

48.59658.07367.5577.027Feb 6, 2025Apr 17, 2025Jun 26, 2025Sep 4, 2025Nov 13, 2025Jan 22, 2026Apr 2, 2026
Updated 15d ago

Evaluation Results

MethodLinks
2026.04
85.1
2026.04
84.7
2026.04
84.6
2026.04
84.3
2026.04
84.2
2026.04
84.1
2026.04
83.7
2026.04
83.5
2026.04
83.4
2026.04
82.8
2026.04
81.6
2026.04
81.3
2025.02
79.55
2026.04
79.4
2026.04
78.9
2026.04
77.5
2026.04
77.3
2026.04
77.2
2026.04
77.1
2026.04
76.9
2026.04
76.9
2025.02
76.6
2025.02
76.52
2025.02
75.05
2025.02
72.83
2026.02
72.47
2026.02
72
2026.02
71.6
2025.02
71.56
2026.02
71.5
2026.02
71
2025.02
70.72
2025.02
70.59
2026.02
69.24
2025.02
69.1
2026.02
68.15
2026.02
68.01
2026.02
68
2025.02
67.55
2026.02
66.89
2026.02
66.07
2026.02
66.01
2026.02
65.68
2026.04
65.6
2026.02
65.19
2026.02
65
2026.02
65
2026.04
65
2026.02
64.91
2026.04
64.6
2026.02
64.54
2026.04
64.4
2026.02
64.18
2026.02
64
2026.02
63.23
2026.02
62.94
2026.02
62.75
2026.04
62.3
2026.02
61.94
2025.02
61.71
2026.02
61.63
2026.02
61.51
2026.02
61
2026.04
60.7
2026.02
60.61
2026.04
60.6
2026.02
60.57
2026.02
60.18
2026.02
60
2026.02
59.35
2026.02
59.3
2026.02
59.1
2026.02
59
2026.02
59
2026.02
59
2025.02
58.97
2026.02
58.82
2026.04
58.5
2026.04
58.5
2026.02
58.44
2026.04
57.3
2026.02
57
2026.04
56.8
2026.02
56
2026.02
56
2026.02
56
2026.04
55.2
2026.02
55
2026.02
55
2026.02
55
2026.04
54.2
2026.02
54
2026.04
53.2
2026.02
53
2026.04
52.9
2026.02
52
2026.04
51.4
2026.02
51
2026.04
50.5
2026.02
50
Showing 100 of 107 rows