Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Active Visual Search on SAT (synthetic)
Loading...
69.33
Accuracy
SFT
56.8292
60.0746
63.32
66.5654
Dec 15, 2025
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
SFT
Backbone=Qwen2.5-VL-7B...
2025.12
69.33
SFT + RL (Ours)
Backbone=Qwen2.5-VL-7B...
2025.12
69.33
Qwen2.5-VL-7B
Backbone=Qwen2.5-VL-7B
2025.12
59.11
RL
Backbone=Qwen2.5-VL-7B...
2025.12
57.31
Feedback
Search any
task
Search any
task
