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

Step-level correctness prediction on GTA

92.17AUROC

PAIR

74.219678.879883.5488.2002May 18, 2026
Updated 15d ago

Evaluation Results

MethodLinks
2026.05
92.1711.98
2026.05
90.8213.42
2026.05
90.3413.56
2026.05
89.7813.78
2026.05
89.6712.94
2026.05
88.5115.21
2026.05
88.2114.32
2026.05
87.4511.82
2026.05
87.099.54
2026.05
74.91-