Explaining Conditions for Reinforcement Learning Behaviors from Real and Imagined Data

Abstract

The deployment of reinforcement learning (RL) in the real world comes with challenges in calibrating user trust and expectations. As a step toward developing RL systems that are able to communicate their competencies, we present a method of generating human-interpretable abstract behavior models that identify the experiential conditions leading to different task execution strategies and outcomes. Our approach consists of extracting experiential features from state representations, abstracting strategy descriptors from trajectories, and training an interpretable decision tree that identifies the conditions most predictive of different RL behaviors. We demonstrate our method on trajectory data generated from interactions with the environment and on imagined trajectory data that comes from a trained probabilistic world model in a model-based RL setting.

arXiv commentAccepted to the Workshop on Challenges of Real-World RL at NeurIPS 2020

References in corpus (6)

• Playing Atari with Deep Reinforcement Learning
• Challenges of Real-World Reinforcement Learning
• Visual Foresight: Model-Based Deep Reinforcement Learning for Vision-Based Robotic Control
• Benchmarking Model-Based Reinforcement Learning
• Interestingness Elements for Explainable Reinforcement Learning: Understanding Agents' Capabilities and Limitations
• Calibrated Model-Based Deep Reinforcement Learning