Thomas Cowperthwaite
I am an Applied Mathematics PhD student at the University of Cambridge, working on probabilistic machine learning, specifically generative models. My research focuses on Gaussian processes and operator learning with an emphasis on machine learning for scientific applications. I'm supported by the AI4ER CDT and supervised by Dr Henry Moss.
Publications
Feature Space Adaptation for Effortless Gaussian Process Flows
Thomas Cowperthwaite, Louis Sharrock, Lachlan Astfalck, and Henry Moss. arXiv preprint arXiv:2610.11459 (2026).
arXiv ·
Scalable Gaussian Process Flows
Thomas Cowperthwaite, Lachlan Astfalck, Henry Moss, and Louis Sharrock. Workshop on AI for Stochastic Dynamics @ NeurIPS (2026).
PDF ·
Conditioning Gaussian Processes on Almost Anything
Henry Moss*, Lachlan Astfalck*, Thomas Cowperthwaite, Colin Doumont, Sam Willis, Philipp Hennig, Christopher Nemeth, and Andrew Zammit-Mangion. NeurIPS 2026. arXiv preprint arXiv:2605.21041 (2026).
arXiv ·
GPGreen: Learning Linear Operators with Gaussian Processes
Thomas Cowperthwaite and Henry Moss. 1st Workshop on Differentiable Systems and Scientific Machine Learning @ EurIPS (2025).
For the latest record of my work, see my Google Scholar.
News
October 2026
- 📝 Feature Space Adaptation for Effortless Gaussian Process Flows is available as an arXiv preprint.
September 2026
- 📝 Our paper on Scalable Gaussian Process Flows was accepted at the Workshop on AI for Stochastic Dynamics at NeurIPS 2026, Sydney!
- 📝 Our paper on Conditioning Gaussian Processes on Almost Anything was accepted at NeurIPS 2026!
- 🎙️ Presenting Conditioning Gaussian Processes on Almost Anything at the Non-Equilibrium Sampling workshop at Newcastle University.
May 2026
- 📝 Conditioning Gaussian Processes on Almost Anything is available as an arXiv preprint.
Blog
Longer notes and updates now live on the blog.