This is the code repository for the paper Probabilistic Neural Circuits.
Please cite as follows:
@inproceedings{zuidberg2024probabilistic,
title = {Probabilistic Neural Circuits},
author = {Zuidberg Dos Martires, Pedro},
booktitle = {Proceedings of the AAAI Conference on Artificial Intelligence (AAAI)},
year = {2024},
}
git clone git@github.com:pedrozudo/ProbabilisticNeuralCircuits.git
cd ProbabilisticNeuralCircuits
python3 -m venv .venv
source .venv/bin/activate
pip install -r requirements.txt
The experiment for density estimation on MNIST can be run using the following command:
python run.py experiments/GenRC/mnist.json
The other experiments can be run by replacing the mnist.json
file with either fmnist.json
or emnist.json
. Note that the first time you perform an experiment with a dataset this will download the dataset itself.
To run the experiments for the discriminative setting use the following command:
python run.py experiments/GenDisRC/mnist.json