Trustworthy Reinforcement Learning Against Intrinsic Vulnerabilities: Robustness, Safety, And Generalizability

Abstract

A trustworthy reinforcement learning algorithm should be competent in solving challenging real-world problems, including \{robustly\} handling uncertainties, satisfying \{safety\} constraints to avoid catastrophic failures, and \{generalizing\} to unseen scenarios during deployments. This study aims to overview these main perspectives of trustworthy reinforcement learning considering its intrinsic vulnerabilities on robustness, safety, and generalizability. In particular, we give rigorous formulations, categorize corresponding methodologies, and discuss benchmarks for each perspective. Moreover, we provide an outlook section to spur promising future directions with a brief discussion on extrinsic vulnerabilities considering human feedback. We hope this survey could bring together separate threads of studies together in a unified framework and promote the trustworthiness of reinforcement learning.

Stats

• citations0
• S2 citations—
• github stars0
• HF likes0
• heat score0.00
• arxiv keyxu2022trustworthy

Related papers

• New Challenges In Reinforcement Learning: A Survey Of Security And Privacy (2022)10.85
• On Assessing The Safety Of Reinforcement Learning Algorithms Using Formal Methods (2021)0.00
• Robust Deep Reinforcement Learning Through Adversarial Attacks And Training : A Survey (2024)0.00
• On The Robustness Of Safe Reinforcement Learning Under Observational Perturbations (2022)0.00
• Safe Reinforcement Learning With Dual Robustness (2023)8.60
• Optimal Transport Perturbations For Safe Reinforcement Learning With Robustness Guarantees (2023)0.00
• Seeing Is Not Believing: Robust Reinforcement Learning Against Spurious Correlation (2023)0.00
• Certifying Safety In Reinforcement Learning Under Adversarial Perturbation Attacks (2022)0.00