I am a PhD student at Mila working on problems in causal representation learning-- currently focusing on structure discovery in the representation space of text-conditioned generative models, and using such structure to explore different notions of interpretable concepts. I also like software engineering and spend time making my code more visual, efficient, and such that it is possible to be lazy with it later.
I love to express myself through intricate prose and poorly drawn stick-figures. Usually curious about too many things.