Hallucination Evaluation
Benchmarks
Dataset NameSOTA methodMetricTrendResultsLast Updated
13.9CO
6
Apr 13, 2026
2.23Score
6
Feb 26, 2026
52.93CS
6
Feb 26, 2026
53.8Event Omission
6
Feb 26, 2026
699.5Total Score
6
Feb 26, 2026
47.3CRs Accuracy
6
Apr 6, 2026
590MME Score (Zh)
6
Feb 26, 2026
48.3Score (Zh)
6
Feb 26, 2026
170Score (Chinese)
6
Feb 26, 2026
70MME Count (Zh)
6
Feb 26, 2026
190Zh Score
6
Feb 26, 2026
87.07Presence
5
May 19, 2026
64.1Average Score
5
May 19, 2026
55Accuracy
5
May 1, 2026
46.5Object-Relation
5
Apr 15, 2026
62.28Entire Score
5
Apr 15, 2026
0.292Co
5
Mar 6, 2026
25.2Ha (Living Room)
5
Feb 26, 2026
2.4Attribute Hijacking Fidelity
4
Jun 17, 2026
3.64F_d
4
May 7, 2026
17.8Question Pair Accuracy
4
Feb 26, 2026
91.16F1 Score
3
May 15, 2026
93.48F1 Score
3
May 15, 2026
83.73Accuracy
2
May 15, 2026
90.53Accuracy
2
May 15, 2026