Hallucination Detection
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
94.26PR-AUC
16
Jun 18, 2026
64.19Accuracy
16
May 19, 2026
85.3AUC-PR
16
May 7, 2026
84.7AUROC (128 steps)
16
Mar 18, 2026
84.2AUROC (128 Steps)
16
Mar 18, 2026
83.28AUCROC
16
Feb 26, 2026
75F1 Score
16
Feb 26, 2026
0.702AUC
15
Jul 7, 2026
78.7AUROC
15
Jun 2, 2026
0.802AUROC
15
Jun 2, 2026
0.77AUROC
15
Jun 2, 2026
52.8AUROC
15
Jun 2, 2026
0.528AUROC
15
Jun 2, 2026
58.7AUROC
15
Jun 2, 2026
0.571AUROC
15
Jun 2, 2026
0.582AUROC
15
Jun 2, 2026
0.704AUROC
15
Jun 2, 2026
0.67AUROC
15
Jun 2, 2026
0.675AUROC
15
Jun 2, 2026
0.844AUROC
15
Jun 2, 2026
72.9F1 Score
15
Apr 9, 2026
0.894AUROC
15
Feb 26, 2026
87.67AUROC
15
Feb 26, 2026
72.2Dialogue Score
15
Apr 6, 2026
88.3Nonfact Accuracy
15
Feb 26, 2026