James Thornton
Research Scientist at Google DeepMind
United Kingdom
Summary
👤
Senior🎓
Top SchoolJames Thornton is a research scientist with nine years of experience specializing in large-scale multimodal generative models, particularly diffusion and flow-based methods and optimal transport. Currently at Google DeepMind after a recent research scientist role at Apple MLR, he has driven both foundational advances (ICML/NeurIPS/AISTATS publications) and applied work in music and media generation. His PhD from Oxford focused on score-based generative models and sampling, and he brings industrial research experience from NVIDIA and internships bridging OT and generative modeling. Known for improving fast flow models and progressive entropic OT solvers, he combines rigorous theoretical insight with pragmatic model engineering at scale. Less obvious: his background in quantitative finance and teaching informs a rare blend of applied problem framing, clear exposition, and production-minded experimental rigor.
9 years of coding experience
2 years of employment as a software developer
Doctor of Philosophy - PhD Computational Statistics and Machine Learning, Doctor of Philosophy - PhD Computational Statistics and Machine Learning at University of OxfordHymers CollegeMasters Statistics with Mathematics, Masters Statistics with Mathematics at University of WarwickThe Snaith School
Github Skills (47)
markdown10
transport10
automatic-differentiation10
gitpod10
github-pages10
jekyll10
parallel9
generative-model8
cuda8
jit8
multi-gpu8
pytorch7
artificial-intelligence7
gpu7
jax7
Programming languages (4)
Github contributions (5)
JTT94/jtt94.github.io
Jun 2020 - Feb 2025
Github Pages template for academic personal websites, forked from mmistakes/minimal-mistakes
Contributions:2 PRs, 217 pushes, 3 branches in 4 years 8 months
templategithub-pages-templatemmistakesmistakesjekyll
JTT94/nlp_neural_process
May 2019 - Aug 2019
Tensorflow implementation of Neural Process, with BERT for NLP tasks
Contributions:27 commits, 4 PRs, 24 pushes in 2 months
nlpbertnlp-taskstensorflownatural-language-processing
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.