What I also do:
(original photography from National Geographic Travellers)
Hi!
I am a master student in Computer Science at the University of Toronto and the Vector Institute, advised by Prof. Chris J. Maddison and informally supervised by Prof. Daniel Roy.
Deep learning was the reason that I left the industry for research. How refreshingly simple artificial neural network architectures are! Often defined by a few lines of formulas, in contrast to the giant mess of biological brains, yet they perform tasks that were once monopolized by brains. Do the symmetries in these architecture adapt particularly well to the problems we care about? Or, are we simply utilizing the flexibility that arise out of a large number of simple units?
I am particularly interested in how each layer of neural networks modifies the geometry of the input space, and how such changes accumulate as the network goes deep. I am currently collaborating with Mufan Li and Lorenzo Noci on the scaling limits of language models at initialisation.
Before this, I worked as a pricing actuary and advanced analytics professional at Definity Financial, where my job was to derive insights from complex real world data. I also lectured at the University of Waterloo. You can find my full professional history on LinkedIn.
Further back, I studied Pure Mathematics and Statistics at University of Waterloo.
What I also do:
(original photography from National Geographic Travellers)