Hallucination Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
95.8Recall
5
Feb 26, 2026
98.6F1 (Faithful)
5
Feb 26, 2026
71.5Accuracy
5
Feb 26, 2026
8Improvement (%)
5
Feb 26, 2026
0.689AUROC (T=0)
5
Feb 26, 2026
91.4Precision
4
Jul 2, 2026
75Precision
4
Jul 2, 2026
100Precision
4
Jul 2, 2026
100Precision
4
Jul 2, 2026
42.4Macro F1
4
Jun 30, 2026
73AUROC (Global)
4
Jun 16, 2026
77AUROC (Global)
4
Jun 16, 2026
0.82AUROC (Global)
4
Jun 16, 2026
72.9AUROC
4
Jun 4, 2026
75.8AUROC
4
May 26, 2026
78.5AUROC
4
May 26, 2026
87.6AUROC (%)
4
May 26, 2026
89.1AUROC
4
May 26, 2026
95.5AUROC
4
May 26, 2026
96.2AUROC (%)
4
May 26, 2026
98.5AUROC
4
May 26, 2026
0.831F1 Score (E4)
4
May 6, 2026
85.4Accuracy
4
Apr 9, 2026
91MC Accuracy
4
Mar 13, 2026
75.4Accuracy
4
Mar 13, 2026