mlsanityData to 14 Oct 2026

Safe and constrained RL

323 papers2 this weekHigh confidence in this grouping: the papers form one clear topicHigh confidence in this grouping: the papers form one clear topic · About

Keywords: safety, safe, reinforcement, reinforcement learning, policy

Key facts. Safe and constrained RL: 323 papers on arXiv since 2024 (data through 14 Oct 2026). 16 of its 323 mapped papers (5%) cite “A Survey of Constraint Formulations in Safe Reinforcement Learning” (2024, arXiv 2402.02025). 10 new papers 30 Sep–14 Oct.

What’s new

30 Sep – 14 Oct 2026

10 new papers, 30 Sep – 14 Oct; 2 this week.

  1. This week · 2
  2. Safe Reinforcement Learning with Receding-Horizon Safety Critic and Differentiable Action Shields
  3. Certifying Almost-Sure Reachability in Unknown MDPs
  4. Earlier · 8
  5. FAITH: Feasibility-Aware Safety-Filtered RL for High-Dimensional Systems
  6. A Unified Bellman Operator for Safety-Critical Reinforcement Learning
  7. Higher-Order Action Supervision Makes A Strong Policy Class also fits: RL for robot locomotion and manipulation
  8. Safe on Average, Unsafe in the Tail: When Is the Episodic-Cost Tail Controllable?
  9. Reachability-Aware Diffusion Policy Optimization also fits: RL theory and sample complexity
  10. FoSeRL: Formal Sequential Robustness Certification for Reinforcement Learning Policies
  11. CORE-RL: Confidence-Oriented Reliability Evaluation of Black-Box Reinforcement Learning Policies
  12. DODGER: Safety-Guided Reinforcement Learning for Robot Navigation Among Dynamic Obstacles

Foundations

Most cited by papers here

  1. A Survey of Constraint Formulations in Safe Reinforcement Learning2024cited by 16 here
  2. Balance Reward and Safety Optimization for Safe Reinforcement Learning: A Perspective of Gradient Manipulation2024cited by 12 here
  3. ActSafe: Active Exploration with Safety Constraints for Reinforcement Learning2024cited by 11 here
  4. Dynamic Model Predictive Shielding for Provably Safe Reinforcement Learning2024cited by 7 here
  5. Safety Filtering While Training: Improving the Performance and Sample Efficiency of Reinforcement Learning Agents2024cited by 7 here

Lineage

The papers this direction builds on, oldest first.

Each paper on the line cites the one before.

  1. 2024A Survey of Constraint Formulations in Safe Reinforcement Learningcited by 16 here
  2. 2024Safe Reinforcement Learning on the Constraint Manifold: Theory and Applicationscited by 5 here
  3. 2024Enhancing Efficiency of Safe Reinforcement Learning via Sample Manipulationcited by 5 here
21 more it builds on, not in this chain
  1. 2024Off-Policy Primal-Dual Safe Reinforcement Learningcited by 5 here
  2. 2024Leveraging Approximate Model-based Shielding for Probabilistic Safety Guarantees in Continuous Environmentscited by 4 here
  3. 2024Safety Filters for Black-Box Dynamical Systems by Learning Discriminating Hyperplanescited by 4 here
  4. 2024Analyzing Adversarial Inputs in Deep Reinforcement Learningcited by 4 here
  5. 2024Barrier Functions Inspired Reward Shaping for Reinforcement Learningcited by 4 here
  6. 2024Sampling-based Safe Reinforcement Learning for Nonlinear Dynamical Systemscited by 4 here
  7. 2024Constrained Reinforcement Learning with Smoothed Log Barrier Functioncited by 4 here
  8. 2024Reinforcement Learning-based Receding Horizon Control using Adaptive Control Barrier Functions for Safety-Critical Systemscited by 5 here
  9. 2024Learning Control Barrier Functions and their application in Reinforcement Learning: A Surveycited by 4 here
  10. 2024Balance Reward and Safety Optimization for Safe Reinforcement Learning: A Perspective of Gradient Manipulationcited by 12 here
  11. 2024Reinforcement learningcited by 10 here
  12. 2024Dynamic Model Predictive Shielding for Provably Safe Reinforcement Learningcited by 7 here
  13. 2024Formally Verifying Deep Reinforcement Learning Controllers with Lyapunov Barrier Certificatescited by 4 here
  14. 2024A CMDP-within-online framework for Meta-Safe Reinforcement Learningcited by 9 here
  15. 2024Exterior Penalty Policy Optimization with Penalty Metric Network under Constraintscited by 5 here
  16. 2024Gymnasium: A Standard Interface for Reinforcement Learning Environmentscited by 20 here
  17. 2024Deep Reinforcement Learning for Robotics: A Survey of Real-World Successescited by 5 here
  18. 2024ActSafe: Active Exploration with Safety Constraints for Reinforcement Learningcited by 11 here
  19. 2024Safety Filtering While Training: Improving the Performance and Sample Efficiency of Reinforcement Learning Agentscited by 7 here
  20. 2024Embedding Safety into RL: A New Take on Trust Region Methodscited by 5 here
  21. 2025SPiDR: A Simple Approach for Zero-Shot Safety in Sim-to-Real Transfercited by 6 here

The 24 most cited of 268.

Browse

30 of 323, ranked 6–35

  1. 6SPiDR: A Simple Approach for Zero-Shot Safety in Sim-to-Real Transfer20256
  2. 7Off-Policy Primal-Dual Safe Reinforcement Learning20245
  3. 8Reinforcement Learning-based Receding Horizon Control using Adaptive Control Barrier Functions for Safety-Critical Systems20245
  4. 9Safe Reinforcement Learning on the Constraint Manifold: Theory and Applications20245
  5. 10Enhancing Efficiency of Safe Reinforcement Learning via Sample Manipulation20245
  6. 11Exterior Penalty Policy Optimization with Penalty Metric Network under Constraints20245
  7. 12Embedding Safety into RL: A New Take on Trust Region Methods20245
  8. 13Probabilistic Shielding for Safe Reinforcement Learning20254
  9. 14Leveraging Approximate Model-based Shielding for Probabilistic Safety Guarantees in Continuous Environments20244
  10. 15Safety Filters for Black-Box Dynamical Systems by Learning Discriminating Hyperplanes20244
All 30
  1. 16Sampling-based Safe Reinforcement Learning for Nonlinear Dynamical Systems20244
  2. 17Constrained Reinforcement Learning with Smoothed Log Barrier Function20244
  3. 18Learning Control Barrier Functions and their application in Reinforcement Learning: A Survey20244
  4. 19Formally Verifying Deep Reinforcement Learning Controllers with Lyapunov Barrier Certificates20244
  5. 20Safe Reinforcement Learning in Black-Box Environments via Adaptive Shielding20244
  6. 21Safe Exploration via Policy Priors20263
  7. 22SafeOR-Gym: A Benchmark Suite for Safe Reinforcement Learning Algorithms on Practical Operations Research Problems20253
  8. 23Safe Reinforcement Learning using Action Projection: Safeguard the Policy or the Environment?20253
  9. 24CBF-RL: Safety Filtering Reinforcement Learning in Training with Control Barrier Functions20253
  10. 25Adaptive Primal-Dual Method for Safe Reinforcement Learning20243
  11. 26Verification-Guided Shielding for Deep Reinforcement Learning20243
  12. 27MAGICS: Adversarial RL with Minimax Actors Guided by Implicit Critic Stackelberg for Convergent Neural Synthesis of Robot Safety20243
  13. 28Learning-Based Shielding for Safe Autonomy under Unknown Dynamics20243
  14. 29A Dynamic Safety Shield for Safe and Efficient Reinforcement Learning of Navigation Tasks20243
  15. 30Extreme Value Policy Optimization for Safe Reinforcement Learning20262
  16. 31Off-Policy Safe Reinforcement Learning with Constrained Optimistic Exploration20262
  17. 32Probabilistically safe and efficient model-based reinforcement learning20252
  18. 33Anytime Safe Reinforcement Learning20252
  19. 34Provably Optimal Reinforcement Learning under Safety Filtering20252
  20. 35Learning safety critics via a non-contractive binary bellman operator20242