Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Multi-modal Hallucination Evaluation on MMHal-Bench
Loading...
1.83
Attribute Failure Score
PAI
1.8
2.0025
2.205
2.4075
Jun 28, 2026
Attribute Failure Score
Adversarial Failure Score
Affordance Failure Score
Counting Failure Score
Spatial Failure Score
Scene Failure Score
OCR Failure Score
Celebrity Recognition Failure Score
Overall Failure Score
Updated 25d ago
Evaluation Results
Method
Method
Links
Attribute Failure Score
Adversarial Failure Score
Affordance Failure Score
Counting Failure Score
Spatial Failure Score
Scene Failure Score
OCR Failure Score
Celebrity Recognition Failure Score
Overall Failure Score
PAI
Model=LLaVA-1.5-7B
2026.06
1.83
0.75
2.33
2
1.92
2.17
1.25
2.42
1.83
VCD
Model=LLaVA-1.5-7B
2026.06
1.83
0.83
1.83
1.17
1.67
4.17
1.58
2.25
1.92
Greedy
Model=LLaVA-1.5-7B
2026.06
2.25
1.33
2.92
1.75
1.92
3.25
1.58
1.42
2.05
FADE
Model=LLaVA-1.5-7B
2026.06
2.42
1.25
2.83
1.75
2.08
3.25
1.42
1.75
2.09
DAMO
Model=LLaVA-1.5-7B
2026.06
2.58
1.33
3
1.75
1.92
3.42
1.17
1.42
2.07
Feedback
Search any
task
Search any
task