About Me

I am currently at Princeton University studying ways to advance science through the use of machine learning with a particular focus on physics-informed and geometric deep learning. Previously I was at the Australian National University where I performed numerical experiments and helped develop a code to simulate cosmic ray propagation through turbulent plasma. I have previously worked as a data scientist/computational statistician in fields of information geometry and biology.

Education

Current PhD - Princeton
Focussing on machine learning for the natural sciences.
Advisors: Peter Melchior and Romain Teyssier
Expected graduation 2026

2023 Masters - Princeton
Masters of Astrophysical Sciences

2021 Honours - ANU
Thesis: Turbulent diffusion of cosmic rays
Major: Computational astrophysics

2020 Bachelor - Mathematics
Queensland University of Technology
Major: Applied and computational

2020 Bachelor - Science
Queensland University of Technology
Major: Physics

Past lives

I previously worked as an ecologist for Ecotone flora and fauna, where my goal was to help reduce the deforestation of far north Queenland rainforests (Weipa).

Directly prior to coming back to my academic studies, I was studying modern guitar at the West Australian Academy for Performing Arts (WAAPA). I still play these days but in a far more limited capacity.

Research

These days I primarily work on developing novel machine learning applications/tools for science. I am intereted in taking advantage of physical symmetries that exist in nature to improve the iductive biases of neural networks to improve both their predictive, and generalisation power. Previously I have worked in areas of astrophysical fluid dynamics and cosmic ray diffusion where I have been directly involved in the code develpment of numerous tools. My publication list can be found below and at these links.
Google - Scholar NASA - ADS Orc - ID

Publications

—— 2024 ——

Score-matching neural networks for improved multi-band source separation
ML Sampson, P Melchior, C Ward, S Birmingham
Submitted to Astronomy and Computing
Link - Paper Link - Code

—— 2023 ——

Spotting Hallucinations in Inverse Problems with Data-Driven Priors
ML Sampson, P Melchior
ICML workshop ML4Astro -- Oral presentation
Link - Paper Link - Code Link - Talk

Turbulent diffusion of streaming cosmic rays in compressible, partially ionized plasma
ML Sampson, JR Beattie, MR Krumholz, RM Crocker, C Federrath, A Seta
Monthly Notices of the Royal Astronomy Society
Link - Paper

—— 2022 ——

Cosmic ray interstellar propagation tool using Itô Calculus (criptic): software for simultaneous calculation of cosmic ray transport and observational signatures
MR Krumholz, RM Crocker, ML Sampson
Monthly Notices of the Royal Astronomy Society
Link - Paper Link - Code

Ion alfvén velocity fluctuations and implications for the diffusion of streaming cosmic rays
JR Beattie, MR Krumholz, C Federrath, ML Sampson, RM Crocker
Frontiers in Astronomy and Space Sciences
Link - Paper

—— 2019 ——

The impact of pair-instability mass loss on the binary black hole mass distribution
S Stevenson, ML Sampson, J Powell, A Vigna-Gómez, C Neijssel, D Szécsi, I Mandel
The Astrophysical Journal
Link - Paper

Teaching

I am passionate about teaching at all levels, from beginner to advanced material in physics, mathematics and computer science. Throughout my studies, I have always been actively involved in teaching both undergraduate and graduate students, specifically in the fields of mathematics and statistics.

Selected courses taught at Princeton and QUT:
Introductory to Astronomy
Statistics for biology
Physics of the very large
Quantitative analysis
Experimental physics

Contact

I'm always happy to talk about astrophysics, machine learning and really anything science related.

Email
matt.sampson@princeton.edu
Address: Peyton Hall
4 Ivy Lane, Princeton

Public profile links: