Yu Guo
Research Scientist at Meta
Palo Alto, California, United States
Summary
🤩
Rockstar🎓
Top SchoolYu Guo is a research scientist at Meta with five years of professional experience bridging deep learning research and low-level systems work. He holds a Ph.D. in Electrical Engineering and an MS in Statistics from Stanford, and his background spans graduate research roles at Stanford and University of Alberta. Practically-minded in production ML, he has contributed robust test coverage and bug fixes to PyTorch and implemented backend compiler optimizations for Triton, including a custom dequantize IR instruction to speed quantized workloads. Based in Palo Alto, he combines strong statistical and engineering foundations to tackle numerical stability and advanced indexing edge cases that often trip up large-scale ML systems. Notably, his open-source work shows a consistent focus on making high-performance tensor tooling both faster and more reliable.
5 years of coding experience
7 years of employment as a software developer
Master's Degree, Electrical and Computer Engineering, Master's Degree, Electrical and Computer Engineering at University of AlbertaMaster of Science - MS, Statistics, Master of Science - MS, Statistics at Stanford UniversityBachelor, Information Engineering, Bachelor, Information Engineering at Zhejiang University
Github Skills (23)
pytorch10
debug10
c-language10
python10
llvm10
testing10
compiler-design10
tensor10
cprogramming-language10
test-automation10
numerical9
machine-learning9
autograd9
numerical-methods9
numerics9
Programming languages (6)
Github contributions (5)
pytorch/pytorch
Apr 2021 - Nov 2022
Tensors and Dynamic neural networks in Python with strong GPU acceleration
Role in this project:
QA Engineer / Test Automation Engineer
QA Engineer / Test Automation EngineerContributions:21 reviews, 48 commits, 42 PRs in 1 year 7 months
Contributions summary:Yu primarily contributed to improving the robustness of the PyTorch library through bug fixes and test additions. Their work focused on identifying and resolving issues related to tensor indexing, empty tensor handling, and ensuring correct behavior of operations like `gather` and `arange`. The commits demonstrate a commitment to thorough testing and addressing potential edge cases within the PyTorch framework, specifically related to advanced indexing and numerical stability.
pythongpu-accelerationdeep-learninggpunumpy
triton-lang/triton
Oct 2022 - Oct 2022
Development repository for the Triton language and compiler
Role in this project:Backend Engineer
Backend EngineerContributions:5 reviews, 1 commit, 5 PRs in 1 day
Contributions summary:Yu primarily contributed to the Triton language and compiler, focusing on low-level optimizations and backend code generation. They introduced a special-purpose `dequantize` instruction within the IR, optimizing for quantized workloads. The user also worked on build system improvements and added environment variables for library paths. Furthermore, they fixed testing scripts and made modifications to existing code.
compilerprogramming-languagecode-generationtriton
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.