Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Reward Prediction on UI-TARS-7B trajectories

96.5Accuracy

PRORE

83.70887.02990.3593.671Sep 26, 2025
Updated 3d ago

Evaluation Results

MethodLinks
2025.09
96.566.7
2025.09
93.922.2
2025.09
9320
2025.09
9320
2025.09
9320
2025.09
89.514.3
2025.09
86.811.7
2025.09
84.225
2025.09
84.210