This repository contains the code for my paper "Emergence of direction-selective retinal cell types in task-optimized deep learning models". All code is writen in Python and deep learning models are coded in PyTorch.
This was my first project using deep learning to model a neural circuit. I don't believe that the code would still work, but in fairness, there are a few clever tricks that someone might find useful. Check out the retina_model.py script to see how I used exec functions to flexibly create different combinations of retinal circuits.
In my current projects, I take a lot of direction from The Good Research Code Handbook to ensure the readability and longevity of my projects. I had a lot of fun working on this project, and I learned a lot, but this code is more nostalgic than anything else.