I'm a computational neuroscientist, research software engineer, applied mathematician, and reformed theoretical physicist.
This website features a curated collection of my research, software, and other work.
I'm currently a research software engineer at Columbia University's Zuckerman Institute, Center for Theoretical Neuroscience.
Curated open-source projects. Full list at github.com/jmrfox.
End-to-end framework for robust multi-view and single-view animal pose estimation, with transformer architectures and temporal context for challenging recordings.
Browser-based workflow for pose projects: label data, train and evaluate models, and run inference on new videos—including multi-view setups.
Publications, posters, and talks.
Arbor simulations of barn-owl toric spines — large, hole-containing spines on space-specific neurons in ICx. The pipeline turns an EM mesh into a cable model, attaches a cylindrical sink, drives synapses from mapped axons, and analyzes compartment voltages.
Side projects, experiments, and toys.
My kid years were spent in Washington state, around Seattle, often slopping through a forest undergrowth, or tripping about a misty tideland, chasing and cataloguing critters, interrogating the innocent flora.
I received a Bachelor's of Science in Physics and Mathematics from Seattle University in 2016. At SU, I contributed to research projects in quantum chromodynamics and computational neuroscience, and I also served as a teaching assistant for undergraduate physics, astronomy, and mathematics courses.
In 2016, I entered the joint PhD program in Computational Science in the Computational Science Research Center at San Diego State University and Claremont Graduate University in California. During my PhD, I was supported by the USDOE Office of Science Graduate Fellowship, working in tandem with the theoretical nuclear physics group at Lawrence Livermore National Laboratory. My dissertation focused on data-driven methods for theoretical nuclear physics: deep learning models for nuclear scattering data libraries, and uncertainty quantification for HPC-scale calculations in the nuclear shell model.
After graduation, I joined the theoretical physics group at Argonne National Laboratory as a postdoctoral research associate. There, I implemented neural network-based improvements for HPC-scale quantum Monte Carlo simulations.
While at Argonne, I decided I wanted to pivot my research focus from physics into computational neuroscience. In 2024, I joined the lab of Jose Peña in the Department of Neuroscience at Albert Einstein College of Medicine in New York City as a postdoctoral research associate. Jose's lab investigates the neural mechanisms of sound localization in the barn owl, and I developed a robust computational pipeline for high-resolution biophysical simulation of auditory neurons in the barn owl's sound-localization system.
I am now a research software engineer in Liam Paninski's lab at Columbia University's Zuckerman Institute, home of the Center for Theoretical Neuroscience. There, I develop software tools for behavioral neuroscience research in collaboration with theorists and experimentalists. Presently, I'm working on Lightning Pose, an end-to-end framework for robust multi-view pose estimation.