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

AntVelocity on Safety Gym

0.93Normalized Reward

Reference policy (πref)

0.21240.39870.5850.7713May 20, 2026
Updated 13d ago

Evaluation Results

MethodLinks
2026.05
0.938.04
2026.05
0.930.13
2026.05
0.90.66
2026.05
0.814.23
2026.05
0.790.5
2026.05
0.240.1