Share your thoughts, 1 month free Claude Pro on us
See more
Home
/
Benchmarks
Instruction-guided image editing preference prediction on AURORA-Bench
Loading...
63.62
Accuracy
EDITREWARD
29.1024
38.0637
47.025
55.9863
Sep 30, 2025
Accuracy
Updated 1mo ago
Evaluation Results
Method
Method
Links
Accuracy
EDITREWARD
Backbone=MiMo-VL-7B-SFT
2025.09
63.62
EDITREWARD
Backbone=Qwen2.5-VL-7B
2025.09
59.5
ADIEE
Method Category=Open-S...
2025.09
55.56
GPT-4o
Method Category=Propri...
2025.09
50.81
Gemini-2.5-Flash
Method Category=Propri...
2025.09
47.63
GPT-5
Method Category=Propri...
2025.09
47.27
Gemini-2.0-Flash
Method Category=Propri...
2025.09
44.31
Qwen2.5-VL-7B-Inst
Method Category=Open-S...
2025.09
38.62
Qwen2.5-VL-32B-Inst
Method Category=Open-S...
2025.09
37.06
Random
Method Category=Baseline
2025.09
33.43
Qwen2.5-VL-3B-Inst
Method Category=Open-S...
2025.09
30.69
MiMo-VL-7B-SFT-2508
Method Category=Open-S...
2025.09
30.43
Feedback
Search any
task
Search any
task