Personal Website

Yubei Chen

Reveal Contact Info

Postdoctoral Scholar

Center for Data Science, NYU; Meta AI (FAIR)

Olshausen Lab

Current Research

News: I’m applying for a faculty position this year!
New paper: Minimalistic Unsupervised Learning with the Sparse Manifold Transform [arXiv]

I have long been obsessed with unsupervised representation learning. For over a decade, I’ve sought the principles governing unsupervised representation learning in human brains and machine learning models. While the empirical advancements in deep learning are tremendous, our understanding and true knowledge about the representation are yet limited. To move further, I believe it’s necessary to invest my time in discovering the principles of the elementary mathematical structures of natural signals. There are not many of these structures, and each of them is not necessarily complex. But, when these simple elements entangle with each other, things quickly exceed our imagination’s ability. My job is to identify and factorize them clearly, concisely, and completely. I am incredibly grateful to receive much help from many great minds along my journey, and I wish to produce some original ideas to entertain my audience.

I like Taoism in Chinese culture, and I think it’s been quite inspiring during my journey voyaging in signal spaces. My favorite philosophical point of view is that “为学日益，为道日损.” This is my light in the darkness.

Background

Postdoctoral Associate at NYU Center for Data Science and Meta AI (FAIR), Oct. 2020 – Present
             Advisor: Professor Yann LeCun
Postdoctoral Associate, Redwood Center for Theoretical Neuroscience, Jan. 2020 – Oct. 2020
             Advisor: Professor Bruno Olshausen
Ph.D. in Electrical Engineering and Computer Science, UC Berkeley, Dec. 2019
             with Designated Emphasis in Communication, Computation, and Statistics
             Advisor: Professor Bruno Olshausen
             Thesis: The Sparse Manifold Transform and Unsupervised Learning for Signal Representation
M.S. in Electrical Engineering and Computer Science, UC Berkeley, Dec. 2018
M.A. in Mathematics, UC Berkeley, May 2015
B.E. in Electrical Engineering, Tsinghua University, Beijing, Jul. 2012

I received my bachelor’s degree from the Electrical Engineering department at Tsinghua University, Beijing, in 2012. Then, I joined the EECS department and Berkeley AI Research (BAIR) at UC Berkeley to pursue my Ph.D. study on unsupervised learning and generative models, advised by Professor Bruno Olshausen. Along the way, I received my M.S. degree in EECS and M.A. degree in mathematics at Berkeley. In 2012, I was awarded the NSF GRFP fellowship. In 2019, I got my Ph.D. from the EECS department. In late 2020, I started to work with Yann LeCun at Meta AI (FAIR) and the Center for Data Science at NYU as a postdoctoral scholar, where I continued to work on unsupervised representation learning. I also serve as a reviewer for NeurIPS, ICLR, ICML, AAAI, CVPR, ECCV, Neural Computation, etc. When I’m away from my research, I enjoy practicing mixed martial art (MMA), and I’m an active member of the UC Berkeley Judo Team.