Safety Representations for Safer Policy Learning

TL;DR

This paper introduces a novel method for reinforcement learning that learns safety representations to enable safer and more efficient exploration in safety-critical environments, outperforming existing constrained exploration techniques.

Contribution

We propose a method that explicitly learns state-conditioned safety representations to improve safe exploration without excessive caution, enhancing policy learning efficiency.

Findings

Significantly improves task performance in safety-critical environments

Reduces constraint violations during training

Balances exploration and safety effectively

Abstract

Reinforcement learning algorithms typically necessitate extensive exploration of the state space to find optimal policies. However, in safety-critical applications, the risks associated with such exploration can lead to catastrophic consequences. Existing safe exploration methods attempt to mitigate this by imposing constraints, which often result in overly conservative behaviours and inefficient learning. Heavy penalties for early constraint violations can trap agents in local optima, deterring exploration of risky yet high-reward regions of the state space. To address this, we introduce a method that explicitly learns state-conditioned safety representations. By augmenting the state features with these safety representations, our approach naturally encourages safer exploration without being excessively cautious, resulting in more efficient and safer policy learning in safety-critical scenarios. Empirical evaluations across diverse environments show that our method significantly improves task performance while reducing constraint violations during training, underscoring its effectiveness in balancing exploration with safety.