What is AutoRL?

Reinforcement learning (RL) is a simple, yet powerful paradigm for training intelligent agents to perform a given task. To do so, RL agents interact with the world they exist in. Guided by a reward signal, RL agents learn in a trial-and-error fashion. That is, RL agents follow a policy, observe if the policy was good or bad and, depending on this outcome, update their policy to get better at a given task.

The simplicity of this paradigm raises the expectations that RL should be applicable in a wide variety of problem domains. However, in practice, it is well known that existing RL algorithms are brittle, require attention to minute implementation details and are sensitive to the experimental setup in general. As a result, RL has not found wide-spread adoption for many real-world tasks.

AutoRL attempts overcome these issues by applying techniques from AutoML and Meta-Learning to RL, making it more efficient, more reliable and more general. This site aims to provide a hub for information on the state of AutoRL from our own groups in Freiburg and Hannover and beyond.

Until we fully flesh out this site, see https://www.automl.org/automated-reinforcement-learning/ for more information.