GENE.jax

This codebase is the official repository to the paper: Searching Search Spaces: Meta-evolving a Geometric Encoding for Neural Networks.

It can also work as python/JAX library to encode Neural Networks using various encodings, like the GENE¹ encoding.

Abstract

In evolutionary policy search, neural networks are usually represented using a direct mapping: each gene encodes one network weight. Indirect encoding methods, where each gene can encode for multiple weights, shorten the genome to reduce the dimensions of the search space and better exploit permutations and symmetries. The Geometric Encoding for Neural network Evolution (GENE) introduced an indirect encoding where the weight of a connection is computed as the (pseudo-)distance between the two linked neurons, leading to a genome size growing linearly with the number of genes instead of quadratically in direct encoding. However GENE still relies on hand -crafted distance functions with no prior optimization. Here we show that better performing distance functions can be found for GENE using Cartesian Genetic Programming (CGP) in a meta-evolution approach, hence optimizing the encoding to create a search space that is easier to exploit. We show that GENE with a learned function can outperform both direct encoding and the hand-crafted distances, generalizing on unseen problems, and we study how the encoding impacts neural network properties.

Citing the work

If you use gene.jax or just want to cite the article, please use the following:

@inproceedings{10612026,
	title     = {Searching Search Spaces: Meta-evolving a Geometric Encoding for Neural Networks},
	author    = {Kunze, Tarek and Templier, Paul and Wilson, Dennis G.},
	year      = 2024,
	booktitle = {2024 IEEE Congress on Evolutionary Computation (CEC)},
	pages     = {1--8},
	doi       = {10.1109/CEC60901.2024.10612026},
	keywords  = {Neurons;Genomics;Genetic programming;Evolutionary computation;Encoding;Bioinformatics;Biological neural networks;evolution strategies;genetic programming;meta-evolution;encoding;neural networks;reinforcement learning;policy search}
}

Footnotes

1. A geometric encoding for neural network evolution - GECCO 21 ↩