modular_rnn

A small package for quick prototyping of multi-region (or modular) recurrent neural network (RNN) models on neuroscience tasks.

Installation

Install in an existing environment:

git clone https://github.com/BeNeuroLab/modular_rnn.git
cd modular_rnn
pip install -e .

Example use

For examples see notebooks modeling the classic center-out reaching task and the "uncertainty task" from Dekleva et al. 2016.

To just quickly play around with the examples, you can use uv:

1. Install uv following the instructions, which on Linux or MacOS is just:

   curl -LsSf https://astral.sh/uv/install.sh | sh

2. uv run --extra examples jupyter lab center_out_example.ipynb