Jason Chou
Independent Researcher at Self-employed
Seattle, Washington, United States
Summary
🤩
Rockstar🎓
Top SchoolJason Chou is an independent researcher and software engineer with a decade of experience building production ML systems and contributing to open-source AI tooling. He holds a PhD in Applied Physics and a Summa Cum Laude B.S.E. in Computer Science, and has shipped applied ML features at Airbnb (fraud prediction, CLIP/OPT-based entity extraction and photo-embedding pipelines) after earlier roles at Google and startups. His recent research outputs include a first-authored ECCV 2024 workshop paper on CLIP geometry and a solo ICLR 2024 blog post on masked language modeling with ALiBi and a CLAP head, bridging academic insight with production practice. Comfortable across Python, TensorFlow, and ML data platforms, he has a track record of turning research ideas into deployable pipelines and contributing to projects like open_clip and torchvision. Colocated in Seattle, he blends rigorous scientific training with hands-on engineering, often surfacing subtle geometric and architectural insights that improve real-world model behavior.
10 years of coding experience
13 years of employment as a software developer
Doctor of Philosophy (Ph.D.), Applied Physics, Doctor of Philosophy (Ph.D.), Applied Physics at University of Michigan - Rackham Graduate SchoolB.S.E., Summa Cum Laude, Computer Science, B.S.E., Summa Cum Laude, Computer Science at University of Michigan
Chinese, Japanese, Portuguese, Spanish
Github Skills (61)
pytorch10
language-model10
python10
attention-mechanism10
machine-learning10
numpy10
deep-learning10
tensorflow10
clip10
gpu10
autograd10
neural-network10
computer-vision10
visual-question-answering10
tensor10
Programming languages (5)
Github contributions (5)
EIFY/open_clip
Jul 2023 - Mar 2026
OpenCLIP fork for "Embedding Geometries of Contrastive Language-Image Pre-Training"
Contributions:1 review, 14 PRs, 87 pushes in 2 years 8 months
EIFY/fairseq
Aug 2022 - Feb 2024
Facebook AI Research Sequence-to-Sequence Toolkit written in Python.
Contributions:83 PRs, 68 pushes, 51 branches in 1 year 6 months
aipythonsequence-to-sequence
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.