Harry Winston Sullivan

Harry Winston Sullivan

PhD Student and Bayesian Bro

Sarupria & Tadmor Groups, University of Minnesota

About

Interests
  • Generative Modeling
  • Gaussian Processes
  • Molecular Simulation
  • Statistical Mechanics
  • Diffusion Models
Education

  • B.Sc. Major in Computational Physics and Minor in Computer Science, 2019 - 2023

    University of Utah

  • PhD in Chemical Engineering, 2024 - Present

    University of Minnesota

Publications

C. Zeng, Harry Winston Sullivan, T. Egg, M. M. Martirossyan, P. Höllmer, J. Jin, R. G. Hennig, A. Roitberg, S. Martiniani, E. B. Tadmor, M. Liu (2026). MolCrystalFlow: Molecular Crystal Structure Prediction via Flow Matching. arXiv Preprint.

PDF Cite DOI

Harry Winston Sullivan, Brennon L. Shanks, M. Cervenka, Michael P. Hoepfner (2025). Physics-Informed Gaussian Process Inference of Liquid Structure from Scattering Data. Journal of Physical Chemistry B.

PDF Cite DOI

Brennon L. Shanks, Harry Winston Sullivan, Pavel Jungwirth, Michael P. Hoepfner (2025). Experimental Evidence of Quantum Drude Oscillator Behavior in Liquids Revealed with Machine Learning Assisted Iterative Boltzmann Inversion. Journal of Chemical Physics.

PDF Cite Code DOI

Brennon L. Shanks, Harry Winston Sullivan, Michael P. Hoepfner (2024). Bayesian Analysis Reveals the Key to Extracting Pair Potentials from Neutron Scattering Data. Journal of Physical Chemistry Letters.

PDF Cite Code DOI

Brennon L. Shanks, Harry Winston Sullivan, Abdur R. Shazed, Michael P. Hoepfner (2024). Accelerated Bayesian Inference for Molecular Simulations using Local Gaussian Process Surrogate Models. Journal of Chemical Theory and Computation.

PDF Cite Code DOI

My Projects

Enter at your own risk…

*
All Physics Machine Learning
Path Integral Molecular Dynamics Tutorial
A tutorial on the theory and implementation of quantum simulation of many body systems.
PDF
Path Integral Molecular Dynamics Tutorial
MNIST Variational Autoencoder
A simple variational autoencoder trained on MNIST. Theory and code included.
MNIST Variational Autoencoder
Clustering Algorithm Demo
This was a class project I made for CS4964 Computational Geometry. It walks through the theory and code behind 4 popular clustering algorithms.
Clustering Algorithm Demo
SOP For Performig BLR in Python
A standard operating procedure for performing Bayesian linear regresssion, in python, from scratch.
PDF
SOP For Performig BLR in Python
Functional Analysis Backing of Quantum Mechanics
These are my typed notes on the mathematics of quantum mechanics. These follow the lectures by Frederic Schuller.
PDF
Functional Analysis Backing of Quantum Mechanics
Varational Fluid Mechanics
Some ramblings of mine while trying to study the principle of least action in terms of fluid mechanics.
PDF
Varational Fluid Mechanics
A Gaussian Process Classifier Made From Scratch, in PyTorch
This explains how to create a GP Classifier with all the theory and code in PyTorch
A Gaussian Process Classifier Made From Scratch, in PyTorch
The Use Of DFT in Catalyst Development
A paper for a course I did in catalyst development and characterization. The class was wholy untheoretical and utterly useless for any computational development. I scratched the itch by making this.
PDF Slides
The Use Of DFT in Catalyst Development