Ta-chu Kao
Research Scientist at Meta
California, United States
Summary
🤩
Rockstar🎓
Top SchoolTa-chu Kao is a research scientist based in California with nine years of experience at the intersection of machine learning, computational neuroscience, and brain–machine interfaces. Currently at Meta working on EMG neuromotor interfaces, he combines deep academic training from Cambridge (PhD) and Oxford (MPhysPhil) with hands-on research roles across industry and academia, including a postdoc at UCL and internships at Reality Labs. He has a strong numerical and algorithmic background, evidenced by significant open-source contributions to Owl (OCaml scientific computing), where he improved automatic differentiation for matrix factorizations and added advanced numerical routines useful for ML and scientific computing. Ta-chu’s work blends theoretical rigor with practical implementation—repairing reverse-mode gradients and fixing broadcasting bugs—to make cutting-edge methods robust and production-ready. Known for moving between low-level numerical detail and high-level neuroscience questions, he brings a rare combination of physics, math, and software craftsmanship to neurotechnology research.
8 years of coding experience
1 year of employment as a software developer
MPhys Phil, Physics and Philosophy, MPhys Phil, Physics and Philosophy at University of OxfordDoctor of Philosophy - PhD, Doctor of Philosophy - PhD at University of Cambridge
Chinese, English, Chinese
Github Skills (11)
numerics10
mat10
automatic-differentiation10
matrix10
numerical10
numerical-methods10
numeric10
scientific-computing10
linear-algebra10
ocaml10
machine-learning8
Programming languages (9)
Github contributions (5)
owlbarn/owl
Jan 2019 - Apr 2021
Owl - OCaml Scientific Computing @ https://ocaml.xyz
Role in this project:
Back-end Developer & Data Scientist
Back-end Developer & Data ScientistContributions:3 reviews, 80 commits, 58 PRs in 2 years 3 months
Contributions summary:Ta-chu primarily contributed to the development and improvement of the automatic differentiation features within the Owl library, focusing on numerical calculations and scientific computing. Their work involved implementing and fixing reverse-mode gradients for various matrix operations, including the QR decomposition, SVD, and Cholesky decomposition. Additionally, they added functions for logdet, bessel functions, and discrete-time Lyapunov equations and addressed associated broadcasting bugs to improve the overall functionality of the automatic differentiation framework. The user's contributions appear to enhance the library's capabilities for machine learning and scientific computing applications.
lapackndarrayautomatic-differentiationpde-solvermatrix
tachukao/ocaml-mujoco
Nov 2021 - Dec 2021
Contributions:64 commits, 48 pushes, 1 branch in 13 days
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.