Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Image Captioning Hallucination Evaluation on MSCOCO CHAIR (val)
Loading...
54.1
CHAIR_S
Qwen3.5-9B
16.14
25.995
35.85
45.705
May 20, 2026
CHAIR_S
CHAIR_I
Average Caption Length
Updated 13d ago
Evaluation Results
Method
Method
Links
CHAIR_S
CHAIR_I
Average Caption Length
Qwen3.5-9B
Backbone=Qwen3.5-9B
2026.05
54.1
10.55
326.47
ILVAD
Backbone=Qwen3.5-9B
2026.05
50.6
9.49
317.98
GLM-4.1V-9B
Backbone=GLM-4.1V-9B
2026.05
26.2
6.36
145.73
Qwen2.5-VL-7B
Backbone=Qwen2.5-VL-7B
2026.05
24.6
7.13
144.48
ILVAD
Backbone=Qwen2.5-VL-7B
2026.05
21.6
6.79
137.09
ILVAD
Backbone=GLM-4.1V-9B
2026.05
17.6
4.18
136.48
Feedback
Search any
task
Search any
task