A Novel Approach to Curiosity and Explainable Reinforcement Learning via Interpretable Sub-Goals
This work addresses the problem of making reinforcement learning more efficient and interpretable for researchers and practitioners, though it appears incremental as it builds on existing curiosity and subgoal methods.
The paper tackled the challenges of sparse rewards and lack of explainability in reinforcement learning by developing an agent that uses a GAN-based curiosity model and subgoal generation to decompose tasks into interpretable steps, achieving improved performance on procedurally-generated tasks with stochastic transitions compared to state-of-the-art methods.
Two key challenges within Reinforcement Learning involve improving (a) agent learning within environments with sparse extrinsic rewards and (b) the explainability of agent actions. We describe a curious subgoal focused agent to address both these challenges. We use a novel method for curiosity produced from a Generative Adversarial Network (GAN) based model of environment transitions that is robust to stochastic environment transitions. Additionally, we use a subgoal generating network to guide navigation. The explainability of the agent's behavior is increased by decomposing complex tasks into a sequence of interpretable subgoals that do not require any manual design. We show that this method also enables the agent to solve challenging procedurally-generated tasks that contain stochastic transitions above other state-of-the-art methods.