List your company

BallRun on Bullet Gym

0.95Normalized RewardReference policy (πref)
0.03480.27240.510.7476May 20, 2026

Evaluation results

MethodLinks
Reference policy (πref)2026.050.953.88
PREFINE2026.050.410.41
BC2026.050.320.97
PPL2026.050.221.99
SafeDICE2026.050.211.71
CPQ2026.050.070.28