A learning gap between neuroscience and reinforcement learning
Machine Learning
2021-05-05 v3 Artificial Intelligence
Abstract
Historically, artificial intelligence has drawn much inspiration from neuroscience to fuel advances in the field. However, current progress in reinforcement learning is largely focused on benchmark problems that fail to capture many of the aspects that are of interest in neuroscience today. We illustrate this point by extending a T-maze task from neuroscience for use with reinforcement learning algorithms, and show that state-of-the-art algorithms are not capable of solving this problem. Finally, we point out where insights from neuroscience could help explain some of the issues encountered.
Cite
@article{arxiv.2104.10995,
title = {A learning gap between neuroscience and reinforcement learning},
author = {Samuel T. Wauthier and Pietro Mazzaglia and Ozan Çatal and Cedric De Boom and Tim Verbelen and Bart Dhoedt},
journal= {arXiv preprint arXiv:2104.10995},
year = {2021}
}
Comments
Accepted as a workshop paper at BRAIN2AI @ ICLR2021