Software

Starting during my PhD and continuing during my work at the Flatiron Institute Center for Computational Neuroscience, I have worked as a developer on several neuroscience-related open source software packages. In rough chronological order, they are:

• pyrtools (core developer, maintainer): tools for multi-scale image processing. pyrtools is a python 3 / numpy port of Eero Simoncelli's matlabPyrTools. Several of its components also have differentiable implementations in plenoptic, below.

• plenoptic (core developer, maintainer): a python library for model-based synthesis of perceptual stimuli, which enables researchers to better understand how their models process images. Powered by pytorch

• CCN template: not an actual python package, but this template is an attempt to make and document all the many decisions that go into writing open-source python packages: structure, documentation, testing, etc.

• nemos (developer): NEural MOdelS, a statistical modeling framework for systems neuroscience, powdered by JAX. It streamlines the process of creating and selecting models, through a collection of easy-to-use methods for feature design. The core of nemos includes GPU-accelerated, well-tested implementations of standard statistical models, currently focusing on the Generalized Linear Model (GLM).