Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Confidence Estimation on VSR (test)
Loading...
67.4
AUROC
Vision-based confidence estimation framework
48.576
53.463
58.35
63.237
Jan 14, 2026
AUROC
Precision
Recall
Coverage @ 60%
Updated 4d ago
Evaluation Results
Method
Method
Links
AUROC
Precision
Recall
Coverage @ 60%
Vision-based confidence estimation framework
VLM=BLIP-2
2026.01
67.4
76.9
26.1
61.9
Vision-based confidence estimation framework
VLM=CLIP
2026.01
58.3
60.7
70.4
54.5
Geometric Only
VLM=CLIP
2026.01
50.9
50.9
55.6
4.2
Khan et al.
VLM=BLIP-2
2026.01
50.3
51.6
100
3.2
Khan et al.
VLM=CLIP
2026.01
50.2
50
1.8
12.2
Geometric Only
VLM=BLIP-2
2026.01
49.3
48.4
90.8
27.6
Feedback
Search any
task
Search any
task