A Python package for creating sparse, randomly-connected, recurrent neural networks, training them using an "innate trajectory" approach, and performing associated experiments.
This package is written in Python 3. To get started, install Python 3 on your computer (we recommend the Conda package manager), then clone this repo onto your computer.
This package depends on:
- matplotlib
- numpy
- scipy
- tqdm
To install them, you can use pip or conda. For pip, type
pip install matplotlib
For conda, type
conda install matplotlib
Repeat this process for each package in the list above.
The file test_main.py is a script that sets experimental parameters and performs the experiment. To run this, set the repo as your current directory and type
python test_main.py
The output figures will be saved as a PDF that will be placed into the figs subdirectory of your repo.
- Michael Seay - mikejseay
This project is licensed under the GNU General Public License v3.0 - see the LICENSE file for details
This code would not be possible without
- Laje, R., & Buonomano, D. V. (2013). Robust timing and motor patterns by taming chaos in recurrent neural networks. Nature Neuroscience, 16(7), 925–933. https://doi.org/10.1038/nn.3405
- Sussillo, D., & Abbott, L. F. (2009). Generating Coherent Patterns of Activity from Chaotic Neural Networks. Neuron, 63(4), 544–557. https://doi.org/10.1016/j.neuron.2009.07.018
This code is the product of work carried out in the group of Dean Buonomano at the University of California Los Angeles. If you find our code helpful to your work, consider citing us in your publications:
- Laje, R., & Buonomano, D. V. (2013). Robust timing and motor patterns by taming chaos in recurrent neural networks. Nature Neuroscience, 16(7), 925–933. https://doi.org/10.1038/nn.3405