Visual Rationalizations in Deep Reinforcement Learning for Atari Games
This addresses the need for explainability in AI systems, particularly for users or developers of reinforcement learning models, but it is incremental as it builds on existing visualization techniques for black-box agents.
The authors tackled the problem of opaque decision-making in deep reinforcement learning agents for Atari games by proposing a method to visualize the evidence behind actions, aiming to make the process more transparent without specifying concrete performance improvements.
Due to the capability of deep learning to perform well in high dimensional problems, deep reinforcement learning agents perform well in challenging tasks such as Atari 2600 games. However, clearly explaining why a certain action is taken by the agent can be as important as the decision itself. Deep reinforcement learning models, as other deep learning models, tend to be opaque in their decision-making process. In this work, we propose to make deep reinforcement learning more transparent by visualizing the evidence on which the agent bases its decision. In this work, we emphasize the importance of producing a justification for an observed action, which could be applied to a black-box decision agent.