Boyuan Chen
Research Scientist at OpenAI
Cambridge, Massachusetts, United States
Summary
👤
Senior🎓
Top SchoolBoyuan Chen is a research scientist at OpenAI with eight years of experience at the intersection of machine learning, robotics, and multimodal models. Trained as a PhD student at MIT EECS and with research stints at Berkeley, DeepMind, and Google X, he has led work on world models, reinforcement learning, and spatial/visual alignment for large language models. At OpenAI he is a core member of teams building GPT image and video generation systems and contributes to RL and world-model training. He is also an active open-source contributor to influential projects such as pytorch/vision and MIT manipulation course materials, improving robustness of image transforms and hands-on robotics exercises. Known for bridging theory and practice, he often ships reproducible code and tests that catch subtle data-handling bugs. Based in Cambridge, MA, he combines strong academic rigor with production-scale ML engineering.
8 years of coding experience
3 years of employment as a software developer
Doctor of Philosophy - PhD, Computer Science, Doctor of Philosophy - PhD, Computer Science at Massachusetts Institute of TechnologyBachelor, Computer Science, Math, Bachelor, Computer Science, Math at University of California - Berkeley
Github Skills (15)
pytorch10
machine-learning10
computer-vision10
deep-learning10
tensorflow10
python10
image-processing10
numpy10
testing9
operation9
jupyter-notebook9
robotics9
tensorrt9
tensor9
reinforcement-learning7
Programming languages (8)
Github contributions (5)
RussTedrake/manipulation
Sep 2022 - Nov 2022
Course notes for MIT manipulation class
Role in this project:
ML Engineer
ML EngineerContributions:1 review, 7 commits, 9 PRs in 1 month
Contributions summary:Boyuan primarily contributes to exercises related to robotic manipulation and deep learning, focusing on computer vision and deep learning concepts. They implemented an exercise involving contrastive loss and made significant modifications to existing notebooks, including normal estimation and policy gradient. Furthermore, the user addressed typos and minor bugs within the codebase, demonstrating a focus on code quality and exercise completion.
manipulationnotespybulletmitcourse-notes
pytorch/vision
Jun 2019 - Aug 2019
Datasets, Transforms and Models specific to Computer Vision
Role in this project:ML Engineer
ML EngineerContributions:5 commits, 7 PRs, 9 comments in 2 months
Contributions summary:Boyuan primarily focused on improving the functionality and correctness of image processing transforms within the `pytorch/vision` repository. They fixed bugs related to how input tensors are handled, ensuring that image data wasn't modified in-place. The user also implemented tests to verify the correct behavior of these transforms in different contexts, particularly for detection models. These contributions directly improve the usability and reliability of image transformations for computer vision tasks within the repository.
pytorchvisiondeep-learningdatasetcomputer-vision
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.