Beyond Hard Constraints: Budget-Conditioned Reachability For Safe Offline Reinforcement Learning

• URL: http://arxiv.org/abs/2603.22292v2
• Date: Tue, 31 Mar 2026 13:32:58 GMT
• Title: Beyond Hard Constraints: Budget-Conditioned Reachability For Safe Offline Reinforcement Learning
• Abstract summary: We show how a safety reachability set enforces safety constraints without unstable min/max or Lagrangian optimization.<n>We also show how this set enforces safety constraints without unstable min/max or Lagrangian optimization.<n>Our method matches or outperforms state of the art baselines while maintaining safety.
• Score: 13.674504740468159
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Sequential decision making using Markov Decision Process underpins many realworld applications. Both model-based and model free methods have achieved strong results in these settings. However, real-world tasks must balance reward maximization with safety constraints, often conflicting objectives, that can lead to unstable min/max, adversarial optimization. A promising alternative is safety reachability analysis, which precomputes a forward-invariant safe state, action set, ensuring that an agent starting inside this set remains safe indefinitely. Yet, most reachability based methods address only hard safety constraints, and little work extends reachability to cumulative cost constraints. To address this, first, we define a safetyconditioned reachability set that decouples reward maximization from cumulative safety cost constraints. Second, we show how this set enforces safety constraints without unstable min/max or Lagrangian optimization, yielding a novel offline safe RL algorithm that learns a safe policy from a fixed dataset without environment interaction. Finally, experiments on standard offline safe RL benchmarks, and a real world maritime navigation task demonstrate that our method matches or outperforms state of the art baselines while maintaining safety.

Related papers

• Selective Safety Steering via Value-Filtered Decoding [54.87935112120107]
  Large language models (LLMs) are trained to align with human values, but their generations may still violate safety constraints.<n>Existing decoding-time steering methods often intervene unnecessarily, modifying generations that would have been safe under the base model.<n>We propose a new test-time steering method designed to reduce such unnecessary interventions while improving the safety of unsafe responses.
  arXiv  Detail & Related papers  (2026-05-14T12:13:08Z)
• Cooptimizing Safety and Performance Using Safety Value-Constrained Model Predictive Control [14.944399584182017]
  We propose a safety-value function-based terminal constraint that enforces membership in a control-invariant safe set at the end of each planning horizon.<n>This formulation enables real-time synthesis of trajectories that are both high-performing and provably safe.<n>We validate the proposed framework through simulation and hardware experiments on a Flexiv Rizon 10s manipulator.
  arXiv  Detail & Related papers  (2026-04-26T20:10:03Z)
• Safe Flow Q-Learning: Offline Safe Reinforcement Learning with Reachability-Based Flow Policies [0.764671395172401]
  offline safe reinforcement learning seeks reward-maximizing policies from static datasets under strict safety constraints.<n>We propose Safe Flow Q-Learning (SafeFQL), which extends FQL to safe offline RL by combining a Hamilton--Jacobi reachability-inspired safety value function with an efficient one-step flow policy.<n>Across boat navigation, and Safety Gymnasium MuJoCo tasks, SafeFQL matches or exceeds prior offline safe RL performance while substantially reducing constraint violations.
  arXiv  Detail & Related papers  (2026-03-16T11:28:47Z)
• BarrierSteer: LLM Safety via Learning Barrier Steering [83.12893815611052]
  BarrierSteer is a novel framework that formalizes safety by embedding learned non-linear safety constraints directly into the model's latent representation space.<n>We show that BarrierSteer substantially reduces adversarial success rates, decreases unsafe generations, and outperforms existing methods.
  arXiv  Detail & Related papers  (2026-02-23T18:19:46Z)
• PoSafeNet: Safe Learning with Poset-Structured Neural Nets [49.854863600271614]
  existing approaches often enforce multiple safety constraints uniformly or via fixed priority orders, leading to infeasibility and brittle behavior.<n>We formalize this setting as poset-structured safety, modeling safety constraints as a partially ordered set and treating safety composition as a structural property of the policy class.<n>Building on this formulation, we propose PoSafeNet, a differentiable neural safety layer that enforces safety via sequential closed-form projection.
  arXiv  Detail & Related papers  (2026-01-29T22:03:32Z)
• Latent Safety-Constrained Policy Approach for Safe Offline Reinforcement Learning [7.888219789657414]
  In safe offline reinforcement learning (RL), the objective is to develop a policy that maximizes cumulative rewards while strictly adhering to safety constraints.<n>We address these issues with a novel approach that begins by learning a conservatively safe policy through the use of Conditional Variational Autoencoders.<n>We frame this as a Constrained Reward-Return Maximization problem, wherein the policy aims to optimize rewards while complying with the inferred latent safety constraints.
  arXiv  Detail & Related papers  (2024-12-11T22:00:07Z)
• Uniformly Safe RL with Objective Suppression for Multi-Constraint Safety-Critical Applications [73.58451824894568]
  The widely adopted CMDP model constrains the risks in expectation, which makes room for dangerous behaviors in long-tail states. In safety-critical domains, such behaviors could lead to disastrous outcomes. We propose Objective Suppression, a novel method that adaptively suppresses the task reward maximizing objectives according to a safety critic.
  arXiv  Detail & Related papers  (2024-02-23T23:22:06Z)
• Safe Offline Reinforcement Learning with Feasibility-Guided Diffusion Model [23.93820548551533]
  We propose FISOR (FeasIbility-guided Safe Offline RL), which allows safety constraint adherence, reward, and offline policy learning. In FISOR, the translated optimal policy for the translated optimization problem can be derived in a special form of weighted behavior cloning. We show that FISOR is the only method that can guarantee safety satisfaction in all tasks, while achieving top returns in most tasks.
  arXiv  Detail & Related papers  (2024-01-19T14:05:09Z)
• Iterative Reachability Estimation for Safe Reinforcement Learning [23.942701020636882]
  We propose a new framework, Reachability Estimation for Safe Policy Optimization (RESPO), for safety-constrained reinforcement learning (RL) environments. In the feasible set where there exist violation-free policies, we optimize for rewards while maintaining persistent safety. We evaluate the proposed methods on a diverse suite of safe RL environments from Safety Gym, PyBullet, and MuJoCo.
  arXiv  Detail & Related papers  (2023-09-24T02:36:42Z)
• SaFormer: A Conditional Sequence Modeling Approach to Offline Safe Reinforcement Learning [64.33956692265419]
  offline safe RL is of great practical relevance for deploying agents in real-world applications. We present a novel offline safe RL approach referred to as SaFormer.
  arXiv  Detail & Related papers  (2023-01-28T13:57:01Z)
• Enforcing Hard Constraints with Soft Barriers: Safe Reinforcement Learning in Unknown Stochastic Environments [84.3830478851369]
  We propose a safe reinforcement learning approach that can jointly learn the environment and optimize the control policy. Our approach can effectively enforce hard safety constraints and significantly outperform CMDP-based baseline methods in system safe rate measured via simulations.
  arXiv  Detail & Related papers  (2022-09-29T20:49:25Z)

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.