Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Vision-Language Reasoning on MMStar cleaned
Loading...
77.59
Score
Jigsaw + CARE
72.9412
74.1481
75.355
76.5619
Dec 16, 2025
Score
Updated 3d ago
Evaluation Results
Method
Method
Links
Score
Jigsaw + CARE
Training Environment=J...
2025.12
77.59
Qwen2.5 VL
Backbone=Qwen 2.5 VL 7...
2025.12
76.61
Mix + CL + CARE
Training Environment=M...
2025.12
76.34
GRPO-CARE
Backbone=Qwen 2.5 VL 7...
2025.12
75.36
VisualSphinx
Backbone=Qwen 2.5 VL 7...
2025.12
75.27
Jigsaw + CL + CARE
Training Environment=J...
2025.12
74.82
Vision-Zero
Backbone=Qwen 2.5 VL 7...
2025.12
74.46
Jigsaw
Training Environment=J...
2025.12
74.46
Visual Jigsaw
Backbone=Qwen 2.5 VL 7...
2025.12
74.11
ViCrit
Backbone=Qwen 2.5 VL 7...
2025.12
73.12
Feedback
Search any
task
Search any
task