Repo for experiments for the paper: "Investigating the scalability and biological plausibility of the activation relaxation algorithm".
Simply git clone
the repository to your home computer. The main.py
file will run the main model to test whether the frozen feedfoward pass assumption can be relaxed. The cnn.py will run the CNN experiments.
The code is written in [Python 3.x] and uses the following packages:
- [NumPY]
- [PyTorch] version 1.3.1
- [matplotlib] for plotting figures
If you enjoyed the paper or found the code useful, please cite as:
@article{millidge2020investigating,
title={Investigating the Scalability and Biological Plausibility of the Activation Relaxation Algorithm},
author={Millidge, Beren and Tschantz, Alexander and Seth, Anil and Buckley, Christopher L},
journal={arXiv preprint arXiv:2010.06219},
year={2020}
}