Towards Automated Semantic Interpretability in Reinforcement Learning via Vision-Language Models
This addresses the need for transparent and verifiable decision-making in RL for applications requiring interpretability, though it is incremental by combining existing methods.
The paper tackles the problem of achieving semantic interpretability in reinforcement learning by automating feature extraction with vision-language models and training interpretable tree-based policies, resulting in iTRACE matching black-box policy performance while outperforming other interpretable baselines across domains like Atari games and driving.
Semantic interpretability in Reinforcement Learning (RL) enables transparency and verifiability of decision-making. Achieving semantic interpretability in reinforcement learning requires (1) a feature space composed of human-understandable concepts and (2) a policy that is interpretable and verifiable. However, constructing such a feature space has traditionally relied on manual human specification, which often fails to generalize to unseen environments. Moreover, even when interpretable features are available, most reinforcement learning algorithms employ black-box models as policies, thereby hindering transparency. We introduce interpretable Tree-based Reinforcement learning via Automated Concept Extraction (iTRACE), an automated framework that leverages pre-trained vision-language models (VLM) for semantic feature extraction and train a interpretable tree-based model via RL. To address the impracticality of running VLMs in RL loops, we distill their outputs into a lightweight model. By leveraging Vision-Language Models (VLMs) to automate tree-based reinforcement learning, iTRACE loosens the reliance the need for human annotation that is traditionally required by interpretable models. In addition, it addresses key limitations of VLMs alone, such as their lack of grounding in action spaces and their inability to directly optimize policies. We evaluate iTRACE across three domains: Atari games, grid-world navigation, and driving. The results show that iTRACE outperforms other interpretable policy baselines and matches the performance of black-box policies on the same interpretable feature space.