# Runtime-Safety-Guided Policy Repair

arXiv: https://arxiv.org/abs/2008.07667 · HTML: https://mlsanity.com/paper/2008.07667/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Aug 2020.

## Built on by (most cited first, 6 of 6)

- [Neural Network Repair with Reachability Analysis](https://mlsanity.com/paper/2108.04214/) (2021)
- [Rethinking Inverse Reinforcement Learning: from Data Alignment to Task Alignment](https://mlsanity.com/paper/2410.23680/) (2024)
- [Neural Control and Certificate Repair via Runtime Monitoring](https://mlsanity.com/paper/2412.12996/) (2024)
- [Follow the STARs: Dynamic ω-Regular Shielding of Learned Policies](https://mlsanity.com/paper/2505.14689/) (2025)
- [PAGAR: Taming Reward Misalignment in Inverse Reinforcement Learning-Based Imitation Learning with Protagonist Antagonist Guided Adversarial Reward](https://mlsanity.com/paper/2306.01731/) (2023)
- [MORTAR: A Model-based Runtime Action Repair Framework for AI-enabled Cyber-Physical Systems](https://mlsanity.com/paper/2408.03892/) (2024)

## Abstract (arXiv)

We study the problem of policy repair for learning-based control policies in safety-critical settings. We consider an architecture where a high-performance learning-based control policy (e.g. one trained as a neural network) is paired with a model-based safety controller. The safety controller is endowed with the abilities to predict whether the trained policy will lead the system to an unsafe state, and take over control when necessary. While this architecture can provide added safety assurances, intermittent and frequent switching between the trained policy and the safety controller can result in undesirable behaviors and reduced performance. We propose to reduce or even eliminate control switching by `repairing' the trained policy based on runtime data produced by the safety controller in a way that deviates minimally from the original policy. The key idea behind our approach is the formulation of a trajectory optimization problem that allows the joint reasoning of policy update and safety constraints. Experimental results demonstrate that our approach is effective even when the system model in the safety controller is unknown and only approximated.