Yihe Dong
Research Engineer at Princeton University
New Jersey, United States
Summary
👤
Senior🎓
Top SchoolYihe Dong is a math-driven research engineer with 11 years' experience applying geometric deep learning and robust statistical methods to real-world ML problems, currently working on language model research at Princeton. He has held research engineering roles at Google and Microsoft where he developed scalable algorithms for vision-language grounding, efficient optimal transport, private ML, and large-scale nearest neighbor search, with multiple NeurIPS/ICML publications. Equally comfortable in code and theory, he has contributed to high-profile repositories such as google-research and implemented reproducible experiments and ranking algorithms. Yihe’s background in pure mathematics informs his practical focus on provable, efficient solutions for graph representation learning and robust statistics, and his personal projects and website reflect a taste for clear, reproducible research.
11 years of coding experience
7 years of employment as a software developer
Mathematics, Mathematics at University of Wisconsin-MadisonDual enrollment while in high school, Dual enrollment while in high school at University of Georgia - Franklin College of Arts and SciencesMathematics, Mathematics at Princeton University
Github Skills (4)
machine-learning10
tensorflow10
python10
bash8
Programming languages (6)
Github contributions (5)
google-research/google-research
Dec 2022 - Dec 2022
Google Research
Role in this project:
ML Engineer
ML EngineerContributions:1 commit in 1 day
Contributions summary:Yihe contributed to the implementation and refinement of machine learning models within the Google Research repository. Their commits focused on modifying scripts, particularly `run_exp.sh` and `KNF/evaluation.py`, to replicate experimental results and add artifacts related to the paper. The changes include modifications to model parameters, dataset configurations, and evaluation metrics, indicating an active role in model training and result verification. Furthermore, the user committed source scripts, specifically for AdaRank, suggesting a focus on developing and evaluating ranking algorithms within the broader research context.
googlemachine-learningai
twistedcubic/learn-to-hash
Oct 2019 - Jan 2021
Contributions:9 commits, 9 pushes, 1 branch in 1 year 2 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.