Hi! I’m a second-year computer science PhD student at Stanford University, where I work as a member of the Stanford AI Lab. I’m advised by Chelsea Finn and Jiajun Wu and affiliated with the StatsML and SVL groups. My research and education is generously supported by a Sequoia Capital Stanford Graduate Fellowship. I’m a proud Taiwanese-Canadian.
As you might be able to tell from my publications, I have a broad research interest in AI. However, I try to ground my interests in the grand challenge of endowing robots with key qualities of intelligent behavior. Towards this end, I enjoy thinking about a diverse array of topics, including out-of-distribution generalization, vision, reinforcement learning, meta-learning, meat-learning, planning, self-supervised learning, Bayesian inference, and communication.
Previously, I majored in robotics as an Engineering Science undergraduate at the University of Toronto. During this time, I did research at the Vector Institute with Roger Grosse and Dan Roy. I’ve also spent time at Google Brain with Shane Gu, Berkeley AI Research with Sergey Levine, and the Max Planck Institute for Software Systems with Rupak Majumdar. My first research experiences were in optoelectronics and photonics under Joyce Poon at the University of Toronto and Ming C. Wu at UC Berkeley.
Instrumenting manipulators with hand-centric sensing facilitates out-of-distribution generalization.
Marrying Hamiltonian Monte Carlo with annealed importance sampling results in an unbiased estimate of marginal likelihood that supports pathwise derivatives. Surprisingly, stochastic gradient annealed importance sampling turns out to be inconsistent.
State-of-the-art nonvacuous PAC-Bayes generalization bounds for neural network image classifiers.
A meta-learning agent develops an adaptive curriculum of visuomotor tasks by deep clustering its own trajectories.
A simple yet effective unsupervised meta-learning pipeline for image classification pre‑training.
Efficient, correct-by-construction controller synthesis via on-demand abstraction construction and adaptive spatiotemporal granularity.
Foundation models are pre-trained, self-supervised, large-scale, multi-modality models that exhibit emergent functionalities and homogenize deep learning models.
Controller synthesis for safety specifications with on-demand abstraction construction.
Efficient correct-by-construction controller synthesis via adaptive spatiotemporal granularity.
Wrapping germanium photodetectors around silicon waveguides results in devices with improved optoelectronic properties.