Provably Optimal Reinforcement Learning under Safety Filtering

• URL: http://arxiv.org/abs/2510.18082v1
• Date: Mon, 20 Oct 2025 20:20:10 GMT
• Title: Provably Optimal Reinforcement Learning under Safety Filtering
• Abstract summary: We show that enforcing safety with a permissive safety filter does not degrade inherent performance.<n>We formalize safety with a safety-critical decision process (SC-MDP), which requires categorical, rather than high-probability, avoidance of catastrophic failure states.<n>Our main theorem establishes that (i) learning in the filtered MDP is safe categorically, (ii) standard RL convergence carries over to the filtered MDP, and (iii) any policy that is optimal in the filtered MDP.
• Score: 6.197103843323546
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Recent advances in reinforcement learning (RL) enable its use on increasingly complex tasks, but the lack of formal safety guarantees still limits its application in safety-critical settings. A common practical approach is to augment the RL policy with a safety filter that overrides unsafe actions to prevent failures during both training and deployment. However, safety filtering is often perceived as sacrificing performance and hindering the learning process. We show that this perceived safety-performance tradeoff is not inherent and prove, for the first time, that enforcing safety with a sufficiently permissive safety filter does not degrade asymptotic performance. We formalize RL safety with a safety-critical Markov decision process (SC-MDP), which requires categorical, rather than high-probability, avoidance of catastrophic failure states. Additionally, we define an associated filtered MDP in which all actions result in safe effects, thanks to a safety filter that is considered to be a part of the environment. Our main theorem establishes that (i) learning in the filtered MDP is safe categorically, (ii) standard RL convergence carries over to the filtered MDP, and (iii) any policy that is optimal in the filtered MDP-when executed through the same filter-achieves the same asymptotic return as the best safe policy in the SC-MDP, yielding a complete separation between safety enforcement and performance optimization. We validate the theory on Safety Gymnasium with representative tasks and constraints, observing zero violations during training and final performance matching or exceeding unfiltered baselines. Together, these results shed light on a long-standing question in safety-filtered learning and provide a simple, principled recipe for safe RL: train and deploy RL policies with the most permissive safety filter that is available.

Related papers

• Safe Control using Learned Safety Filters and Adaptive Conformal Inference [0.8921166277011348]
  Safety filters have been shown to be effective tools to ensure the safety of control systems with unsafe nominal policies.<n>This paper presents Adaptive Conformal Filtering (ACoFi), a method that combines learned Hamilton-Jacobi reachability-based safety filters with adaptive conformal inference.<n>We show that ACoFi guarantees that the rate of incorrectly quantifying uncertainty in the predicted safety of the nominal policy is upperally bounded by a user-defined parameter.
  arXiv  Detail & Related papers  (2026-04-20T16:34:49Z)
• Safe Reinforcement Learning via Recovery-based Shielding with Gaussian Process Dynamics Models [57.006252510102506]
  Reinforcement learning (RL) is a powerful framework for optimal decision-making and control but often lacks provable guarantees for safety-critical applications.<n>We introduce a novel recovery-based shielding framework that enables safe RL with a provable safety lower bound for unknown and non-linear continuous dynamical systems.
  arXiv  Detail & Related papers  (2026-02-12T22:03:35Z)
• RSafe: Incentivizing proactive reasoning to build robust and adaptive LLM safeguards [55.76285458905577]
  Large Language Models (LLMs) continue to exhibit vulnerabilities despite deliberate safety alignment efforts.<n>To safeguard against the risk of policy-violating content, system-level moderation via external guard models has emerged as a prevalent mitigation strategy.<n>We propose RSafe, an adaptive reasoning-based safeguard that conducts guided safety reasoning to provide robust protection within the scope of specified safety policies.
  arXiv  Detail & Related papers  (2025-06-09T13:20:04Z)
• A Provable Approach for End-to-End Safe Reinforcement Learning [17.17447653795906]
  A longstanding goal in safe reinforcement learning (RL) is to ensure the safety of a policy throughout the entire process.<n>We propose a method, called Provably Lifetime Safe RL (PLS), that integrates offline safe RL with safe policy deployment.
  arXiv  Detail & Related papers  (2025-05-28T00:48:20Z)
• Safety Modulation: Enhancing Safety in Reinforcement Learning through Cost-Modulated Rewards [23.15178050525514]
  Safe Reinforcement Learning (Safe RL) aims to train an RL agent to maximize its performance in real-world environments while adhering to safety constraints.<n>We propose a novel safe RL approach called Safety Modulated Policy Optimization (SMPO), which enables safe policy function learning.
  arXiv  Detail & Related papers  (2025-04-03T21:35:22Z)
• Probabilistic Shielding for Safe Reinforcement Learning [51.35559820893218]
  In real-life scenarios, a Reinforcement Learning (RL) agent must often also behave in a safe manner, including at training time.<n>We present a new, scalable method, which enjoys strict formal guarantees for Safe RL.<n>We show that our approach provides a strict formal safety guarantee that the agent stays safe at training and test time.
  arXiv  Detail & Related papers  (2025-03-09T17:54:33Z)
• Safety Filtering While Training: Improving the Performance and Sample Efficiency of Reinforcement Learning Agents [7.55113002732746]
  Reinforcement learning (RL) controllers are flexible and performant but rarely guarantee safety. Safety filters impart hard safety guarantees to RL controllers while maintaining flexibility. We analyze several modifications to incorporating the safety filter in training RL controllers rather than solely applying it during evaluation.
  arXiv  Detail & Related papers  (2024-10-15T15:01:57Z)
• Safety through Permissibility: Shield Construction for Fast and Safe Reinforcement Learning [57.84059344739159]
  "Shielding" is a popular technique to enforce safety inReinforcement Learning (RL) We propose a new permissibility-based framework to deal with safety and shield construction.
  arXiv  Detail & Related papers  (2024-05-29T18:00:21Z)
• Safety Optimized Reinforcement Learning via Multi-Objective Policy Optimization [3.425378723819911]
  Safe reinforcement learning (Safe RL) refers to a class of techniques that aim to prevent RL algorithms from violating constraints. In this paper, a novel model-free Safe RL algorithm, formulated based on the multi-objective policy optimization framework is introduced.
  arXiv  Detail & Related papers  (2024-02-23T08:58:38Z)
• Approximate Model-Based Shielding for Safe Reinforcement Learning [83.55437924143615]
  We propose a principled look-ahead shielding algorithm for verifying the performance of learned RL policies. Our algorithm differs from other shielding approaches in that it does not require prior knowledge of the safety-relevant dynamics of the system. We demonstrate superior performance to other safety-aware approaches on a set of Atari games with state-dependent safety-labels.
  arXiv  Detail & Related papers  (2023-07-27T15:19:45Z)
• Safe Reinforcement Learning via Confidence-Based Filters [78.39359694273575]
  We develop a control-theoretic approach for certifying state safety constraints for nominal policies learned via standard reinforcement learning techniques. We provide formal safety guarantees, and empirically demonstrate the effectiveness of our approach.
  arXiv  Detail & Related papers  (2022-07-04T11:43:23Z)
• Conservative Safety Critics for Exploration [120.73241848565449]
  We study the problem of safe exploration in reinforcement learning (RL) We learn a conservative safety estimate of environment states through a critic. We show that the proposed approach can achieve competitive task performance while incurring significantly lower catastrophic failure rates.
  arXiv  Detail & Related papers  (2020-10-27T17:54:25Z)
• Cautious Reinforcement Learning with Logical Constraints [78.96597639789279]
  An adaptive safe padding forces Reinforcement Learning (RL) to synthesise optimal control policies while ensuring safety during the learning process. Theoretical guarantees are available on the optimality of the synthesised policies and on the convergence of the learning algorithm.
  arXiv  Detail & Related papers  (2020-02-26T00:01:08Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.