Recent Group Interests: Our current focus is on the principles of language modeling. Broadly, we aim to develop principled methodologies to fundamentally understand, interpret, and leverage the mechanics and phenomena underlying LLM training and inference. Our research spans multiple layers, from model design—where we explore whether simplified yet accurate models can enable tractable analysis while retaining key system properties—to both experimental and theoretical investigations. For preliminary work in this area, check this and this. Thus, we particularly welcome individuals who have experience with training and inference of (large) language models and are also driven by conceptual questions about gaining a deeper understanding of what these models do, where they don't, and why.
UBC Under-Graduate Students
I welcome inquiries from prospective research interns. If browsing my website and papers sparks your curiosity, and you have a strong background in coding and/or mathematics, please consider reaching out.
To apply, please share your CV and transcripts. The best way to understand my work style and interests is by taking one of my courses (see the Teaching section). Additionally, I highly recommend completing a basic machine learning course before joining the group.
Graduate Students
Unfortunately, I am not able to respond to all the emails from applicants. I apologize in advance for that and I still make an effort to read all of them. Thank you for reaching out showing your interest.
I welcome MASc/Ph.D. applications from students with strong background in mathematics and interests in deep-learning and machine learning fundamentals. I suggest you read some of my recent research papers to see whether the topics fit your background and interests.
My group is committed to equity, diversity and inclusion and I encourage all interested students to apply or contact me with any questions.