Benjamin Kurt Miller
Summary
🤩
RockstarBenjamin Miller is a research scientist with 11 years of experience at the intersection of machine learning, physics, and computational chemistry, currently working on FAIR Chemistry at Meta. He completed a PhD in Computer Science at the University of Amsterdam after research stints with Max Welling and groups in AI4Science, and has driven ML-for-science projects at Microsoft Research and Meta’s Open Catalyst. Benjamin contributes to e3nn, enhancing mathematical foundations and equivariance tests for neural networks with Euclidean symmetry—work that reflects a deep comfort with both theory and engineering. His background in physics and computational science fuels practical innovations like high-performance molecular data loaders and symmetry-respecting models, and he has a track record of speeding production pipelines and shipping reproducible research. Notably, he blends formal mathematical improvements (spherical harmonic parity operators, SO3 fixes) with hands-on backend and ML engineering to make advanced equivariant architectures more robust and usable.
11 years of coding experience
Github Skills (12)
tensorrt10
neural-network10
pytorch10
machine-learning10
tensor10
tensorflow10
operation10
python10
linear-algebra9
computational-geometry9
data-structure7
data-structures7
Programming languages (9)
Github contributions (5)
e3nn/e3nn
Jul 2019 - Jun 2020
A modular framework for neural networks with Euclidean symmetry
Role in this project:
Back-end Developer & ML Engineer
Back-end Developer & ML EngineerContributions:2 releases, 89 commits, 10 PRs in 11 months
Contributions summary:Benjamin primarily contributed to the `se3nn` project, focusing on modifications related to the mathematical foundations and practical application of neural networks with Euclidean symmetry. Their work involved bug fixes in convolution operations, as well as improvements to the `SO3` and parity functionalities. They added functionality for spherical harmonic parity operators and implemented an equivariance test, demonstrating a focus on both the theoretical aspects and the practical performance of the project.
neural-network
bkmi/cnre
Oct 2022 - Jul 2024
Extending amortized approximate likelihood-ratio estimation / neural ratio estimation to the contrastive case, while staying consistent with the original work.
Contributions:3 releases, 1 PR, 14 pushes in 1 year 9 months
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.