Evolving Neuronal Plasticity Rules using Cartesian Genetic Programming

TL;DR

This paper uses Cartesian genetic programming to evolve biologically plausible synaptic plasticity rules, enabling neural networks to learn tasks like principal component analysis with rules tailored to dataset structures.

Contribution

It introduces a method to automatically discover interpretable plasticity rules for neural networks using evolutionary algorithms, applicable across various learning paradigms.

Findings

Evolved plasticity rules perform competitively with hand-designed solutions.

Dataset properties influence the form of the evolved rules.

New dataset-adapted plasticity rules are discovered.

Abstract

We formulate the search for phenomenological models of synaptic plasticity as an optimization problem. We employ Cartesian genetic programming to evolve biologically plausible human-interpretable plasticity rules that allow a given network to successfully solve tasks from specific task families. While our evolving-to-learn approach can be applied to various learning paradigms, here we illustrate its power by evolving plasticity rules that allow a network to efficiently determine the first principal component of its input distribution. We demonstrate that the evolved rules perform competitively with known hand-designed solutions. We explore how the statistical properties of the datasets used during the evolutionary search influences the form of the plasticity rules and discover new rules which are adapted to the structure of the corresponding datasets.