Leabra.jl

This is a Julia conversion of the old version of the Leabra algorithm written in Matlab by Sergio Verduzco. The original code be found here: https://github.com/emer/cemer/tree/master/Matlab

The initial commits of this are almost an exact conversion, save for implementation of the NXX1 function which is based on the Go code found here: https://github.com/emer/leabra

An example network that can give you an idea of how to use this can be found in example_network.jl

Installation

This code requires Julia 1.8. It was developed with the beta version, but really its only using const for some structs.

It will install the following libraries into the project environment

• Distributions.jl
• LinearAlgebra.jl
• Plots.jl
• Random.jl

Ways to run the example:

• Clone this repo into a directory via git clone https://github.com/jshahbazi/Leabra.jl
• Run cd Leabra.jl
• Run julia --project=. -e "using Pkg; Pkg.instantiate()"
• Run julia --project=. example_network.jl

Or, you can run it from the Julia shell via these steps:

• Run julia --project=.
• From the julia> prompt, run: include("example_network.jl")

Or, just use the example_notebook.ipynb. It has the code from example_network.jl in an easy-to-use format.

Citation

Leabra and all associated ideas can be found here:

O’Reilly, R. C., Munakata, Y., Frank, M. J., Hazy, T. E., and Contributors (2012). Computational Cognitive Neuroscience. Wiki Book, 4th Edition (2020). URL: https://CompCogNeuro.org