about

I am a researcher at OpenAI Frontiers Research, doing capability research with Reiichiro Nakano (WebGPT lead) and Hyeonwoo Noh (GPT-4V lead).

Before OpenAI, I spent some time in academia. I am on leave from the Stanford CS PhD program and spent a wonderful year before at CMU MLD, advised by Prof. Andrej Risteski. My research focused on the mathematical and scientific foundation of deep learning, particularly problems in generative models and distribution shift robustness.

Before that, I graduated Summa Cum Laude, Phi Beta Kappa, Tau Beta Pi, Upsilon Pi Epsilon, College Honors from UCLA with B.S. in Computer Science and B.S. in Applied Mathematics. There, I was very fortunate to work with Prof. Jonathan Kao on robot learning, brain-computer interface, and computational neuroscience, and with Prof. Wei Wang on natural language processing and domain adaptation.

I love to build. At Google X, I built reinforcement learning infrasturcture for the Everyday Robots. I also led the design, implementation, launch and iterations of the cloud serverless ML infrastructure at Percolata.

In a past life, I was a competitive sailor and synthetic bio-hacker before I got indefinitely distracted by the beauty of mathematics, algorithms, and the brain.