# Improving Safety in Reinforcement Learning Using Model-Based Architectures and Human Intervention

arXiv: https://arxiv.org/abs/1903.09328 · HTML: https://mlsanity.com/paper/1903.09328/

## Key facts

- Cited by 8 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 22 Mar 2019.

## Built on by (most cited first, 8 of 8)

- [Look Before You Leap: Safe Model-Based Reinforcement Learning with Human Intervention](https://mlsanity.com/paper/2111.05819/) (2021)
- [Play to Grade: Testing Coding Games as Classifying Markov Decision Process](https://mlsanity.com/paper/2110.14615/) (2021)
- [Towards an Interpretable Hierarchical Agent Framework using Semantic Goals](https://mlsanity.com/paper/2210.08412/) (2022)
- [Safe Reinforcement Learning for Power System Control: A Review](https://mlsanity.com/paper/2407.00681/) (2024)
- [Modeling Risk in Reinforcement Learning: A Literature Mapping](https://mlsanity.com/paper/2312.05231/) (2023)
- [Action Shapley: A Training Data Selection Metric for World Model in Reinforcement Learning](https://mlsanity.com/paper/2601.10905/) (2026)
- [Probabilistic Counterexample Guidance for Safer Reinforcement Learning (Extended Version)](https://mlsanity.com/paper/2307.04927/) (2023)
- [Multiagent Copilot Approach for Shared Autonomy between Human EEG and TD3 Deep Reinforcement Learning](https://mlsanity.com/paper/2312.14458/) (2023)

## Abstract (arXiv)

Recent progress in AI and Reinforcement learning has shown great success in solving complex problems with high dimensional state spaces. However, most of these successes have been primarily in simulated environments where failure is of little or no consequence. Most real-world applications, however, require training solutions that are safe to operate as catastrophic failures are inadmissible especially when there is human interaction involved. Currently, Safe RL systems use human oversight during training and exploration in order to make sure the RL agent does not go into a catastrophic state. These methods require a large amount of human labor and it is very difficult to scale up. We present a hybrid method for reducing the human intervention time by combining model-based approaches and training a supervised learner to improve sample efficiency while also ensuring safety. We evaluate these methods on various grid-world environments using both standard and visual representations and show that our approach achieves better performance in terms of sample efficiency, number of catastrophic states reached as well as overall task performance compared to traditional model-free approaches