Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Visual Reasoning on Hull-Bench
Loading...
67.3
Accuracy
Unsilencing Latent Reasoning
58.9904
61.1477
63.305
65.4623
May 4, 2026
Accuracy
Updated 29d ago
Evaluation Results
Method
Method
Links
Accuracy
Unsilencing Latent Reasoning
Backbone=Qwen2.5VL-7B
2026.05
67.3
LVR
Backbone=Qwen2.5VL-7B
2026.05
66.67
DMLR
Backbone=Qwen2.5VL-7B
2026.05
65.83
Monet
Backbone=Qwen2.5VL-7B
2026.05
65.67
ICoT
Backbone=Qwen2.5VL-7B
2026.05
65.51
Vanilla
Backbone=Qwen2.5VL-7B
2026.05
65.4
LVRRF
Backbone=Qwen2.5VL-7B
2026.05
65.19
CCoT
Backbone=Qwen2.5VL-7B
2026.05
64.88
ICoT
Backbone=Qwen2.5VL-3B
2026.05
64.67
DMLR
Backbone=Qwen2.5VL-3B
2026.05
64.67
CoVT
Backbone=Qwen2.5VL-7B
2026.05
64.46
Unsilencing Latent Reasoning
Backbone=Qwen2.5VL-3B
2026.05
64.35
Vanilla
Backbone=Qwen2.5VL-3B
2026.05
64.14
CCoT
Backbone=Qwen2.5VL-3B
2026.05
64.03
MCoT
Backbone=Qwen2.5VL-3B
2026.05
63.8
MCoT
Backbone=Qwen2.5VL-7B
2026.05
63.62
LVRRF
Backbone=Qwen2.5VL-3B
2026.05
60.99
LVR
Backbone=Qwen2.5VL-3B
2026.05
59.31
Feedback
Search any
task
Search any
task