Search papers, labs, and topics across Lattice.
University of Notre Dame
2
0
3
Safe-RULE can effectively neutralize the impact of data poisoning in offline Safe RL, enhancing safety without the need for retraining.
Safe RL policies, designed to avoid unsafe actions, can be effectively attacked using a novel framework that learns safety constraints from demonstrations and then crafts adversarial perturbations, even without access to the target policy's gradients.
