I am an incoming MSc student in computer science at the University of Toronto, where I am co-advised by Professor Nicolas Papernot. I am also a DeepMind MSc fellow for 2025-2026.

My research focuses at the intersection of ML fairness and security, as motivated by perspectives in AI ethics. That means I view ML fairness as a sort of security for the least powerful and privileged. For example, one of my recent works looks at a phenomenon called model-induced distribution shifts, which is when a model’s outputs contribute to changes in future training sets, biasing newer models. Over time, this can lead to negative impacts to minoritized groups, including their eventual erasure from datasets in extreme cases.

Another of my recent works (with Nick Jia) looked at detecting backdoor attacks. However, unlike many works which assume the defender is a powerful model trainer, our paper works in the setting the defender is a low-resource cloud compute client who needs to ensure their model was not backdoored by a compute provider.

Download my CV.