Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Object Hallucination Detection on MS COCO 2014 (val)
Loading...
90
Accuracy
HaloProbe
48.9928
59.6389
70.285
80.9311
Apr 7, 2026
Accuracy
AUROC
Precision
Recall
F1 Score
Updated 11d ago
Evaluation Results
Method
Method
Links
Accuracy
AUROC
Precision
Recall
F1 Score
HaloProbe
Backbone=LLaVA-1.5-7B
2026.04
90
93.5
92.5
95.8
94.1
DIML
Backbone=LLaVA-1.5-7B
2026.04
84.46
90.19
-
72.34
-
EAZY
Backbone=LLaVA-1.5-7B
2026.04
78.77
-
78.41
83.38
80.82
IC
Backbone=LLaVA-1.5-7B
2026.04
62.56
-
61.93
81.6
70.42
UT
Backbone=LLaVA-1.5-7B
2026.04
50.57
-
53.6
70.62
60.95
Feedback
Search any
task
Search any
task