Wei Wen - Research Scientist at Meta

Wei Wen

Research Scientist at Meta

Bellevue, Washington, United States

Join Prog.AI to see contacts

Summary

🤩

Rockstar

🎓

Top School

Wei Wen is a Research Scientist and tech lead with 11 years of experience building and shipping scalable ML systems, currently leading LLM and VLM development for AR/VR/XR at Meta in Menlo Park. His work spans AutoML, efficient and distributed deep learning, and multimodal models applied to vision, language, and recommender systems, with hands-on expertise from CUDA kernel development to production deployments. He has driven company-wide impact through generative AI and efficiency initiatives, and previously contributed research used in production (TernGrad) while at Facebook AI and open-source extensions to Caffe for sparse/low-rank networks. A Duke PhD, he combines deep theoretical background with practical GPU and platform engineering—an uncommon mix that enables him to shepherd models from research prototypes to device-scale products.

11 years of coding experience

8 years of employment as a software developer

Doctor of Philosophy - PhD, Computer Engineering, Doctor of Philosophy - PhD, Computer Engineering at Duke University

Master's degree, Electrical, Electronics and Communications Engineering, Master's degree, Electrical, Electronics and Communications Engineering at Beihang University

English, Chinese

Github Skills (9)

cuda10

gpu-programming10

machine-learning10

c-language10

caffe10

deep-learning10

cprogramming-language10

mask-rcnn9

faster-rcnn9

Programming languages (6)

C++ShellLuaHTMLJupyter NotebookPython

Github contributions (5)

wenwei202/caffe

Oct 2015 - Mar 2020

Caffe for Sparse and Low-rank Deep Neural Networks

Role in this project:

Back-end Developer

Contributions:497 commits, 1 PR, 475 pushes in 4 years 5 months

Contributions summary:Wei implemented a custom layer (`SparsifyLayer`) in Caffe to sparsify the outputs of a previous layer. This involved writing CUDA kernels for both the forward and backward passes of the new layer, demonstrating expertise in GPU programming. They also added a new parameter to the protobuf definitions to integrate the layer within the Caffe framework. The changes included adding layer-wise display and thresholding functionality.

caffedeep-neural-networkssparsityaccelerationcompression

wenwei202/FastFlexANN

Jun 2015 - Jun 2015

Contributions:33 commits, 32 pushes, 1 branch in 3 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.

Request Free Trial