On the Robustness of Safe Reinforcement Learning under Observational PerturbationsDownload PDF


22 Sept 2022, 12:32 (modified: 08 Nov 2022, 14:43)ICLR 2023 Conference Blind SubmissionReaders: Everyone
Keywords: Safe reinforcement learning, deep reinforcement learning, state robust reinforcement learning
TL;DR: We study the robustness of safe RL under observational perturbations, and propose two effective adversaries and a defense algorithm to increase the agent's safety under attacks.
Abstract: Safe reinforcement learning (RL) trains a policy to maximize the task reward while satisfying safety constraints. While prior works focus on the performance optimality, we find that the optimal solutions of many safe RL problems are not robust and safe against carefully designed observational perturbations. We formally analyze the unique properties of designing effective state adversarial attackers in the safe RL setting. We show that baseline adversarial attack techniques for standard RL tasks are not always effective for safe RL and proposed two new approaches - one maximizes the cost and the other maximizes the reward. One interesting and counter-intuitive finding is that the maximum reward attack is strong, as it can both induce unsafe behaviors and make the attack stealthy by maintaining the reward. We further propose a more effective adversarial training framework for safe RL and evaluate it via comprehensive experiments (video demos are available at: https://sites.google.com/view/robustsaferl/home). This paper provides a pioneer work to investigate the safety and robustness of RL under observational attacks for future safe RL studies.
Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.
No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.
Supplementary Material: zip
Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics
Submission Guidelines: Yes
Please Choose The Closest Area That Your Submission Falls Into: Reinforcement Learning (eg, decision and control, planning, hierarchical RL, robotics)
11 Replies