About Me

  • I am a PhD student at Mila advised by Prof. Sarath Chandar. My research focuses on understanding how deep learning models work (empirically and theoretically) and uncovering their training dynamics. My main insterest is in long-term dependence modelling and memory augmented neural networks, in particular the use of structured memory and how to leverage it for adaptating to distributional shifts.
  • Before this, I spent time in the Bay Area working as a data engineer and recieved bachelor degrees from the University of Waterloo.
  • I am actively seeking motivated students for collaboration. If you are passionate about research and eager to work on innovative projects in deep learning, consider working with me. You will have the opportunity to lead your own research projects or contribute to ongoing studies. To apply for an internship, please fill out this form and directly reach out via email :).
  • My research is supported by NSERC, FRQNT and Hydro-Québec among other institutions and companies.

Recent Works

  • EpiK-Eval: Evaluation for Language Models as Epistemic Models [Link]
  • Promoting Exploration in Memory-Augmented Adam using Critical Momenta [Link]

Contacting Me

  • Please use my e-mail. I check it daily and respond in a timely fashion unless I am traveling or inordinately busy with other things.

Other Information

  • I grew up in a sea-side city in Japan next to a volcano but later moved to Canada. You may have guessed from my name that I have roots in Taiwan.
  • When not at my computer, I read books and cook. I also bike, swim, play badminton and fish quite a lot. During the winter, I also snowboard.