List your company

HalfCheetahVelocity on Safety Gym

99Normalized RewardReference policy (πref)
15.837.45980.6May 20, 2026

Evaluation results

MethodLinks
Reference policy (πref)2026.05998.01
PREFINE2026.05860.15
SafeDICE2026.05700.31
BC2026.05574.1
PPL2026.05280
CPQ2026.05190.71