Advantage Actor-Critic with Reasoner: Explaining the Agent's Behavior from an Exploratory Perspective

Muzhe Guo, Feixu Yu, Tian Lan +1

Reinforcement learning (RL) is a powerful tool for solving complex decision-making problems, but its lack of transparency and interpretability has been a major challenge in domains…