I am a research scientist at OpenAI, where I work on multimodal post-training for the GPT thinking models.
Previously I was at Google DeepMind where I built robot world models. Our work turned video models into simulators for robot policies.
I completed my PhD at Stanford, advised by Chelsea Finn. There, my core research focused on geometric deep learning and nested optimization. I obtained my BS at UC Berkeley, advised by Anca Dragan.