List your company

Safe Reinforcement Learning on Safety-Gymnasium PointPush (OSRL)

0CostCPQ
-3.82421.98847.873.612Jun 8, 2026

Evaluation results

MethodLinks
CPQAttack=Max Cost, Poison Ratio=15%, Unlearning Status=Before2026.060-12
Safe-RULE (CPQ)Attack=Max Cost, Poison Ratio=15%, Unlearning Status=After2026.0600.1
Safe-RULE (CPQ)Attack=Max Reward, Poison Ratio=5%, Unlearning Status=After2026.0601.5
Safe-RULE (BEAR-Lag)Attack=Max Reward, Poison Ratio=15%, Unlearning Status=After2026.0602.9
CPQAttack=Min Reward, Poison Ratio=5%, Unlearning Status=Before2026.060.8-2
Safe-RULE (CPQ)Attack=Max Cost, Poison Ratio=5%, Unlearning Status=After2026.061.41.9
CPQAttack=Min Reward, Poison Ratio=15%, Unlearning Status=Before2026.062-1.2
Safe-RULE (CPQ)Attack=Max Reward, Poison Ratio=15%, Unlearning Status=After2026.062.41.9
Safe-RULE (BEAR-Lag)Attack=Max Cost, Poison Ratio=15%, Unlearning Status=After2026.062.50.7
CPQAttack=Max Cost, Poison Ratio=5%, Unlearning Status=Before2026.063.5-1.7
CPQAttack=Max Reward, Poison Ratio=5%, Unlearning Status=Before2026.063.9-23.6
Safe-RULE (BEAR-Lag)Attack=Min Reward, Poison Ratio=5%, Unlearning Status=After2026.064.21.8
Safe-RULE (COptiDICE)Attack=Max Cost, Poison Ratio=5%, Unlearning Status=After2026.065.81.2
Safe-RULE (COptiDICE)Attack=Max Reward, Poison Ratio=15%, Unlearning Status=After2026.066-2.7
Safe-RULE (CPQ)Attack=Min Reward, Poison Ratio=5%, Unlearning Status=After2026.066.41.4
Safe-RULE (BEAR-Lag)Attack=Max Cost, Poison Ratio=5%, Unlearning Status=After2026.067.42.2
Safe-RULE (COptiDICE)Attack=Min Reward, Poison Ratio=5%, Unlearning Status=After2026.067.51.2
Safe-RULE (BEAR-Lag)Attack=Min Reward, Poison Ratio=15%, Unlearning Status=After2026.067.61.6
Safe-RULE (COptiDICE)Attack=Max Reward, Poison Ratio=5%, Unlearning Status=After2026.068.20.3
Safe-RULE (CPQ)Attack=Min Reward, Poison Ratio=15%, Unlearning Status=After2026.068.51.4
Safe-RULE (BEAR-Lag)Attack=Max Reward, Poison Ratio=5%, Unlearning Status=After2026.068.62.2
Safe-RULE (BCQ-Lag)Attack=Max Reward, Poison Ratio=5%, Unlearning Status=After2026.069.22.9
Safe-RULE (BCQ-Lag)Attack=Max Cost, Poison Ratio=15%, Unlearning Status=After2026.069.62.6
Safe-RULE (BCQ-Lag)Attack=Min Reward, Poison Ratio=15%, Unlearning Status=After2026.069.73.4
Safe-RULE (BCQ-Lag)Attack=Min Reward, Poison Ratio=5%, Unlearning Status=After2026.069.83.8
Safe-RULE (COptiDICE)Attack=Min Reward, Poison Ratio=15%, Unlearning Status=After2026.069.8-1
Safe-RULE (COptiDICE)Attack=Max Cost, Poison Ratio=15%, Unlearning Status=After2026.0610.30.4
Safe-RULE (BCQ-Lag)Attack=Max Reward, Poison Ratio=15%, Unlearning Status=After2026.0610.61.6
CPQAttack=Max Reward, Poison Ratio=15%, Unlearning Status=Before2026.0611.4-1.5
BEAR-LagAttack=Min Reward, Poison Ratio=5%, Unlearning Status=Before2026.0611.7-11.2
BEAR-LagAttack=Max Reward, Poison Ratio=15%, Unlearning Status=Before2026.0611.8-18.2
BEAR-LagAttack=Max Reward, Poison Ratio=5%, Unlearning Status=Before2026.0613.40.4
Safe-RULE (BCQ-Lag)Attack=Max Cost, Poison Ratio=5%, Unlearning Status=After2026.0614.42
BCQ-LagAttack=Max Cost, Poison Ratio=15%, Unlearning Status=Before2026.0614.5-0.8
BCQ-LagAttack=Max Reward, Poison Ratio=15%, Unlearning Status=Before2026.0627.83.2
BEAR-LagAttack=Min Reward, Poison Ratio=15%, Unlearning Status=Before2026.0628.5-10.9
COptiDICEAttack=Max Reward, Poison Ratio=5%, Unlearning Status=Before2026.0628.7-0.1
BEAR-LagAttack=Max Cost, Poison Ratio=5%, Unlearning Status=Before2026.0631.50.6
COptiDICEAttack=Max Cost, Poison Ratio=5%, Unlearning Status=Before2026.0631.51.6
COptiDICEAttack=Min Reward, Poison Ratio=15%, Unlearning Status=Before2026.0634.4-11.6
BCQ-LagAttack=Max Cost, Poison Ratio=5%, Unlearning Status=Before2026.0635.74.3
COptiDICEAttack=Max Cost, Poison Ratio=15%, Unlearning Status=Before2026.0637.70.3
BCQ-LagAttack=Min Reward, Poison Ratio=15%, Unlearning Status=Before2026.0638.9-6.9
BCQ-LagAttack=Max Reward, Poison Ratio=5%, Unlearning Status=Before2026.0640.53.6
COptiDICEAttack=Min Reward, Poison Ratio=5%, Unlearning Status=Before2026.0645.32.2
COptiDICEAttack=Max Reward, Poison Ratio=15%, Unlearning Status=Before2026.0653-0.4
BCQ-LagAttack=Min Reward, Poison Ratio=5%, Unlearning Status=Before2026.0682.83.7
BEAR-LagAttack=Max Cost, Poison Ratio=15%, Unlearning Status=Before2026.0695.6-3.4