Pith. sign in

REVIEW 11 cited by

A Review of Safe Reinforcement Learning: Methods, Theory and Applications

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2205.10330 v5 pith:FJF72XD5 submitted 2022-05-20 cs.AI cs.LG

classification cs.AIcs.LG
keywords safealgorithmsapplicationsproblemsreviewfivefuturelearning
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

Reinforcement Learning (RL) has achieved tremendous success in many complex decision-making tasks. However, safety concerns are raised during deploying RL in real-world applications, leading to a growing demand for safe RL algorithms, such as in autonomous driving and robotics scenarios. While safe control has a long history, the study of safe RL algorithms is still in the early stages. To establish a good foundation for future safe RL research, in this paper, we provide a review of safe RL from the perspectives of methods, theories, and applications. Firstly, we review the progress of safe RL from five dimensions and come up with five crucial problems for safe RL being deployed in real-world applications, coined as "2H3W". Secondly, we analyze the algorithm and theory progress from the perspectives of answering the "2H3W" problems. Particularly, the sample complexity of safe RL algorithms is reviewed and discussed, followed by an introduction to the applications and benchmarks of safe RL algorithms. Finally, we open the discussion of the challenging problems in safe RL, hoping to inspire future research on this thread. To advance the study of safe RL algorithms, we release an open-sourced repository containing the implementations of major safe RL algorithms at the link: https://github.com/chauncygu/Safe-Reinforcement-Learning-Baselines.git.

Discussion (0). Sign in to comment.

Forward citations

Cited by 11 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Adjustment Speed as a Safety Constraint for Nonstationary Reinforcement Learning

    cs.LG 2026-07 conditional novelty 6.0 of 10

    Forecasting context shifts and treating adaptation demand against calibrated recovery capacity as a safety gate reduces transient violations in a nonstationary highway-driving simulator.

  2. Toward Adaptable Multi-Agent Reinforcement Learning: An Assumption-Aware Review

    cs.AI 2025-07 conditional novelty 6.0 of 10

    A survey proposing adaptability as a three-part taxonomy (learning, policy, scenario-driven) for organizing and evaluating MARL under changing conditions.

  3. Optimistic Exploration for Risk-Averse Constrained Reinforcement Learning

    cs.LG 2025-07 conditional novelty 6.0 of 10

    ORAC combines upper-confidence-bound reward exploration with lower-confidence-bound risk-averse cost constraints and adaptive cost weighting to improve exploration in risk-averse constrained RL.

  4. Safe Planning and Policy Optimization via World Model Learning

    cs.AI 2025-06 conditional novelty 6.0 of 10

    SPOWL is a model-based safe RL method that uses a value-equivalent world model, a Lagrangian-trained safe policy, and adaptive planning thresholds to achieve low-cost, high-reward control on SafetyGymnasium tasks.

  5. SafeOR-Gym: A Benchmark Suite for Safe Reinforcement Learning Algorithms on Practical Operations Research Problems

    cs.LG 2025-06 conditional novelty 6.0 of 10

    SafeOR-Gym offers nine constrained OR environments for safe RL and shows that existing algorithms solve some but fail on mixed-integer or nonconvex instances.

  6. Multi-level Certified Defense Against Poisoning Attacks in Offline Reinforcement Learning

    cs.LG 2025-05 conditional novelty 6.0 of 10

    A DP-based certified defense provides lower bounds on expected cumulative reward and per-state action stability for offline RL under transition- and trajectory-level poisoning, with larger certified radii than COPA.

  7. End-to-End Humanoid Robot Safe and Comfortable Locomotion Policy

    cs.RO 2025-08 conditional novelty 5.0 of 10

    An end-to-end humanoid locomotion policy maps raw LiDAR point clouds to motor commands using P3O with CBF-inspired safety costs and comfort rewards, with sim-to-real tests on a Unitree G1.

  8. HCRMP: A LLM-Hinted Contextual Reinforcement Learning Framework for Autonomous Driving

    cs.RO 2025-05 conditional novelty 5.0 of 10

    The HCRMP planner feeds LLM semantic hints into state representation and critic weighting instead of letting the LLM decide actions, reporting better CARLA driving metrics.

  9. Online Learning Control Strategies for Industrial Processes with Application for Loosening and Conditioning

    eess.SY 2025-06 reject novelty 4.0 of 10

    An adaptive Koopman MPC with historical safety constraints is proposed for tobacco conditioning, but the claimed Cpk improvements come from model-generated advisor-mode trajectories, not real closed-loop control.

  10. Closed-Form Robustness Bounds for Second-Order Pruning of Neural Controller Policies

    cs.RO 2025-06 reject novelty 3.0 of 10

    A single-layer Lipschitz bound for pruning is correct, but the paper's additive multi-layer bound is false and the control-safety framing is unsupported.

  11. Thinking Beyond Tokens: From Brain-Inspired Intelligence to Cognitive Foundations for Artificial General Intelligence and its Societal Impact

    cs.AI 2025-07 conditional novelty 2.0 of 10

    A broad survey arguing that AGI requires modular, memory-augmented, embodied architectures rather than scaled-up token prediction, with a brief proposal to decompose intelligence into five components.

Pith tools