Working memory facilitates reward-modulated Hebbian learning in recurrent neural networks

TL;DR

This paper demonstrates that combining a reservoir network with a dynamic working memory enables biologically plausible reward-modulated Hebbian learning to effectively learn complex sequences, matching the performance of more artificial methods.

Contribution

It introduces a biologically plausible learning scheme that integrates working memory with reward-modulated Hebbian learning in reservoir computing.

Findings

The combined network learns complex sequences efficiently.

Reward-modulated Hebbian learning performs comparably to FORCE learning.

The approach offers a biologically plausible alternative for sequence learning.

Abstract

Reservoir computing is a powerful tool to explain how the brain learns temporal sequences, such as movements, but existing learning schemes are either biologically implausible or too inefficient to explain animal performance. We show that a network can learn complicated sequences with a reward-modulated Hebbian learning rule if the network of reservoir neurons is combined with a second network that serves as a dynamic working memory and provides a spatio-temporal backbone signal to the reservoir. In combination with the working memory, reward-modulated Hebbian learning of the readout neurons performs as well as FORCE learning, but with the advantage of a biologically plausible interpretation of both the learning rule and the learning paradigm.