# Accelerating Safe Reinforcement Learning with Constraint-mismatched Policies

arXiv: https://arxiv.org/abs/2006.11645 · HTML: https://mlsanity.com/paper/2006.11645/

## Key facts

- Cited by 11 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 20 Jun 2020.

## Built on by (most cited first, 11 of 11)

- [Safe Reinforcement Learning by Imagining the Near Future](https://mlsanity.com/paper/2202.07789/) (2022)
- [Trustworthy Reinforcement Learning Against Intrinsic Vulnerabilities: Robustness, Safety, and Generalizability](https://mlsanity.com/paper/2209.08025/) (2022)
- [CUP: A Conservative Update Policy Algorithm for Safe Reinforcement Learning](https://mlsanity.com/paper/2202.07565/) (2022)
- [Improving Safety in Deep Reinforcement Learning using Unsupervised Action Planning](https://mlsanity.com/paper/2109.14325/) (2021)
- [Reinforcement Learning by Guided Safe Exploration](https://mlsanity.com/paper/2307.14316/) (2023)
- [Near-optimal Conservative Exploration in Reinforcement Learning under Episode-wise Constraints](https://mlsanity.com/paper/2306.06265/) (2023)
- [Model-Based Safe Reinforcement Learning with Time-Varying State and Control Constraints: An Application to Intelligent Vehicles](https://mlsanity.com/paper/2112.11217/) (2021)
- [RL, but don't do anything I wouldn't do](https://mlsanity.com/paper/2410.06213/) (2024)
- [DeCOM: Decomposed Policy for Constrained Cooperative Multi-Agent Reinforcement Learning](https://mlsanity.com/paper/2111.05670/) (2021)
- [Guided Safe Shooting: model based reinforcement learning with safety constraints](https://mlsanity.com/paper/2206.09743/) (2022)
- [Learning safe, constrained policies via imitation learning: Connection to Probabilistic Inference and a Naive Algorithm](https://mlsanity.com/paper/2507.06780/) (2025)

## Abstract (arXiv)

We consider the problem of reinforcement learning when provided with (1) a baseline control policy and (2) a set of constraints that the learner must satisfy. The baseline policy can arise from demonstration data or a teacher agent and may provide useful cues for learning, but it might also be sub-optimal for the task at hand, and is not guaranteed to satisfy the specified constraints, which might encode safety, fairness or other application-specific requirements. In order to safely learn from baseline policies, we propose an iterative policy optimization algorithm that alternates between maximizing expected return on the task, minimizing distance to the baseline policy, and projecting the policy onto the constraint-satisfying set. We analyze our algorithm theoretically and provide a finite-time convergence guarantee. In our experiments on five different control tasks, our algorithm consistently outperforms several state-of-the-art baselines, achieving 10 times fewer constraint violations and 40% higher reward on average.