Continual Reinforcement Learning with Complex Synapses

TL;DR

This paper demonstrates that integrating complex synaptic models inspired by biology into reinforcement learning agents can significantly reduce catastrophic forgetting, enabling more effective continual learning across tasks and within tasks.

Contribution

The study introduces a biologically inspired synaptic model into reinforcement learning agents to mitigate catastrophic forgetting across multiple timescales.

Findings

Mitigates catastrophic forgetting in reinforcement learning agents.

Enables continual learning across sequential tasks.

Reduces reliance on experience replay databases.

Abstract

Unlike humans, who are capable of continual learning over their lifetimes, artificial neural networks have long been known to suffer from a phenomenon known as catastrophic forgetting, whereby new learning can lead to abrupt erasure of previously acquired knowledge. Whereas in a neural network the parameters are typically modelled as scalar values, an individual synapse in the brain comprises a complex network of interacting biochemical components that evolve at different timescales. In this paper, we show that by equipping tabular and deep reinforcement learning agents with a synaptic model that incorporates this biological complexity (Benna & Fusi, 2016), catastrophic forgetting can be mitigated at multiple timescales. In particular, we find that as well as enabling continual learning across sequential training of two simple tasks, it can also be used to overcome within-task forgetting by reducing the need for an experience replay database.