I am a second-year PhD student at Northeastern University advised by David Bau.

I am interested in understanding the internal representations of large neural networks, and how these machine-learned systems leverage their representations to perform different tasks. My research interests generally include machine learning, applications in computer vision, and natural language processing (NLP) - with an emphasis on interpretability.

Prior to beginning my PhD, I studied Applied and Computational Mathematics (BS) at Brigham Young University (BYU).

News

[January 2024] – Our Function Vectors paper was accepted at ICLR 2024!

[October 2023] – We released a new preprint! – Function Vectors in Large Language Models.
We find that during in-context learning (ICL) language models create internal abstractions of general-purpose functions that are used to trigger a specific functional behavior. We show how to extract a function vector and demonstrate that it can invoke the demonstrated behavior in a variety of contexts. We also investigate what a function vector contains and explore vector algebra for composing function vectors.
[Project Page] [ArXiv] [Code] [Twitter]