Safe RL

RL: Reinforcement Learning

Overview

Safe reinforcement learning aims to maximize task performance while limiting hazards such as collisions, equipment damage, or unsafe medical decisions, including during exploration and deployment. Common approaches formulate constraints as costs in constrained Markov decision processes and use primal-dual or Lagrangian optimization; others add control-barrier-function, Lyapunov, or reachability shields that filter actions, sometimes using learned models to account for uncertainty. Work also addresses offline learning, human preference feedback, temporal-logic specifications, and applications from robotics and driving to healthcare and energy control. A central challenge is balancing strong, reliable safety guarantees with reward, data efficiency, and adaptation to unfamiliar conditions; benchmarks also show that methods vary across tasks and that average costs can hide rare severe failures.

Latest papers

CardsList

239 papers24 papers in the last four weeks, up 9% on the four weeks before. 0.2% of all new papers.

  1. VESPA: Verifiable Symbolic Policy Adaptation for Safe Reinforcement Learning

    Nov 18, 2026Paul König, Sascha Marton, Stefan Lüdtke +1Safe RLFormal Methods for RL

  2. Algorithm Distillation for Safe In-Context Reinforcement Learning

    Nov 16, 2026Minjae Kwon, Amir Moeini, Shangtong Zhang +1In-Context RLSafe RL

  3. Towards Probabilistically Safe Reinforcement Learning: Theory and Deployment on Humanoid and Aerial Robots

    Nov 14, 2026Pol Mestres, Varun Gabbita, Jason H. K. Siu +4Reinforcement LearningSafe RL

  4. High Probability Risk Control for Online Policy Learning

    Nov 9, 2026Yihong Guo, Drew Prinster, Suchi Saria +1Risk-Aware ControlSafe RL

  5. Probably Safe Sim-to-Real Transfer with Lifetime Validity

    Nov 6, 2026Rohan Tangri, Jan-Peter CalliessSim-to-Real TransferSafe RL

  6. Safety-Focused Reinforcement Learning Using Risk-Sensitive Coupled Control

    Nov 6, 2026Jiabin Lin, Shana Moothedath, Vivek S. BorkarSafe RLRL Control

  7. Safe Reinforcement Learning via Probabilistic Option Shielding

    Nov 6, 2026Gian Luca Ascari, Isidoro Tamassia, Ying Jiao +2Safe RLRL Shielding

  8. Accelerating Safe Reinforcement Learning with Massive Parallelism

    Nov 5, 2026Joonyoung Lim, Younghwan YooSafe RLConstrained RL

  9. Provable Zeroth-Order Safe Exploration for Reinforcement Learning from Human Feedback

    Nov 5, 2026Weiqin Chen, Xinmeng Huang, Dongsheng Ding +1Safe RLReinforcement Learning

  10. A Universal Policy Wrapper With Guarantees

    Nov 2, 2026Georgiy Malaniya, Anton Bolychev, Pavel OsinenkoRL ControlSafe RL

  11. Sequential Recovery Updates for Learned Controllers under Successful-Trajectory Preservation

    Nov 2, 2026Zongtan Li, Tianming Liu, Ye ZhaoRobot Failure RecoveryContinual Robot Learning

  12. Beyond Ensemble Disagreement: Predictive STL Monitoring of RL Locomotion under Distribution Shift

    Nov 1, 2026Qurban Ali, Badaruddin Chachar, Maqsood Ali +1Safe RLRobotic Policy Evaluation

  13. Rare Event Constrained Reinforcement Learning with Large Deviations

    Nov 1, 2026Rahul MisraReinforcement LearningConstrained RL

  14. Hi-CoLA: High Confidence Lower Bound Approximation Based Reinforcement Learning for Flex-Route Transit Operation Control

    Oct 31, 2026Qi Kang, Haris N. Koutsopoulos, Hongliang Guo +1RL ControlSafe RL

  15. Structural Dual Adaptation for Safe Reinforcement Learning

    Oct 31, 2026Runsheng Yu, Zhuoyang Song, Zengde Deng +3Primal-Dual RLConstrained RL

  16. DEPLOY-RL: Active Boundary Discovery and Conservative Certification for Deployable Reinforcement Learning in Safety-Critical Continuous Processes

    Oct 30, 2026Yeojin Jang, Minu Baek, Gihun Gil +8Safe RLActive Learning

  17. Safe Multi-Objective Reinforcement Learning: Zero Constraint Violation and Primal-Averaging Convergence

    Oct 29, 2026Chenxi Li, Yao Shu, Siyang Gao +2Multi-Objective Reinforcement LearningReinforcement Learning

  18. CORMA: Constraint-Routed Decision Mamba for Generalizable Multi-Constraint Offline Safe Reinforcement Learning

    Oct 25, 2026Changxu Ni, Feiran You, Yongan Guo +5Safe RLReinforcement Learning

  19. Physics-Informed Graph-Based Safe RL with Constraint-Specific Critics for Community Battery Coordination in Distribution Network

    Oct 24, 2026Chenghao Huang, Yue Yang, Hao WangPower SystemsSafe RL

  20. RBGA: Residual-Budget Guide Arbitration for Offline-to-Online Safe Reinforcement Learning in Dexterous Manipulation

    Oct 24, 2026Zongyou Yu, Ao Zhou, Xiongbo Shi +7Reinforcement LearningSafe RL

  21. SEVI: Safe Evidential Real-World Reinforcement Learning

    Oct 22, 2026Shaekh Mohammad Shithil, Lantao LiuSafe RLRobotic RL

  22. SAFTER: Safety-Aware Filtering Through Explicit Recoverability for Robust Reinforcement Learning under Distribution Shift

    Oct 21, 2026Benedictus C. G. Cinun, Bin HuSafe RLReinforcement Learning

  23. Prescribed Performance Guided Safe Reinforcement Learning for Roundabout Navigation of Autonomous Vehicles

    Oct 20, 2026Wenhan Li, Yangrui Zhang, Hang Zhou +2Autonomous Driving PlanningRL for Autonomous Driving

  24. RECREG: Recovery-based Shielding for Regular Safety Properties in Reinforcement Learning

    Oct 16, 2026Alexander W. Goodall, Francesco BelardinelliRL ShieldingSafe RL

  25. Safe Reinforcement Learning with Receding-Horizon Safety Critic and Differentiable Action Shields

    Oct 14, 2026Siwei Ju, Jan Peters, Oleg ArenzSafe RLSafety Filtering

  26. A Reinforcement Learning Framework for Safety-Aware Haptic Shared Control in Surgical Robotics

    Oct 11, 2026Zihao Zhang, Karl Handwerker, Balint VargaRobotic ControlHuman-Robot Collaboration