Working Memory Spiking RNN Model

Overview

This repository provides the code for the model and analyses presented in this paper:

Kim R. & Sejnowski TJ. Strong inhibitory signaling underlies stable temporal dynamics and working memory in spiking neural networks. Nature Neuroscience (2020).

Preprint available here.

Recurrent Neural Network (RNN) model

For this work, we used the method that we previously developed (reported here) to construct spiking RNNs to perform WM tasks. For more details and access to the code, please refer to this GitHub repository. The trained RNNs analyzed in the study are also available here.

Experimental data

We also analyzed a publicly available dataset collected and generously shared by Dr. Christos Constantinidis's Lab at Wake Forest School of Medicine. Please refer to this website to get acecss and learn more about the dataset.

Analysis

The code for computing neuronal timescales and analyzing both model and experimental data is implemented in MATLAB (tested in 2016a and 2016b).

Model Analysis

All the scripts related to the model analysis are located in the model folder. The folder also contains a README file showing how to use the included scripts.

Experimental Data Analysis

All the scripts related to the experimental data analysis are located in the data folder. The folder also contains a README file showing how to use the included scripts.

Citation

If you use this repo for your research, please cite our work:

@article {Kim_2020,
  author = {Kim, Robert and Sejnowski, Terrence J.},
  title = {Strong inhibitory signaling underlies stable temporal dynamics and working memory in spiking neural networks},
  year = {2020},
  journal = {Nature Neuroscience},
  doi = {https://doi.org/10.1038/s41593-020-00753-w},
}