Heterogeneous Quantization Regularizes Spiking Neural Network Activity

Spiking neural network simulations from Moyal et al. (2025), tested on Ubuntu 22.04.

Installation

If you would like to run, modify, or extend this project:

• Clone the repository:

  git clone https://github.com/cplab/sapinet_regularization
  

• Create a conda virtual environment (optional):

  conda create -n <env_name> python=3.11
  conda activate <env_name>
  

• Install with pip:

  cd sapinet_regularization
  pip install -e .
  

See setup.py for dependency information.

Simulation

• To reproduce the results and plots, run the master script from its own directory:

  cd src/projects
  python simulation.py -multirun yaml/hq2025.yaml
  

• You may modify component, model, and project YAMLs to your liking and run single-use pipelines, e.g.:

  cd src/projects
  python simulation.py -experiment yaml/synthetic.yaml
  

Plots and diagnostic output will be written to results/<run>.

Analysis

To reproduce the statistical analyses:

• Edit the following R file so that run_dir and meta_dir point to the run directory generated in the previous step:

  src/utils/analysis/analysis.r
  

• Run analysis.r and inspect the results generated under analysis/<run>.

You may also directly inspect the raw synthetic datasets and statistical output reported in the manuscript. Both are included in this repository under:

  analysis/SciRep-Final

Note

Sapicore 0.4 is compatible with both Linux and Windows.