I am a PhD student at the University of Cambridge, where I am advised by Ferenc Huszár and David Krueger, and generously funded by a scholarship from Twitter. My research focus is on empirical approaches to understanding how deep learning works, especially at scale. I am also interested in policy considerations for the responsible development of AI.

I completed an MSc at the Université de Montréal and Mila, where I was advised by Laurent Charlin and worked at the intersection of self-supervised learning and deep reinforcement learning. I previously spent a few years working at Airbnb on site performance and anti-fraud initiatives. Prior to that, I did my undergrad in Software Engineering and Computer Science at the University of Waterloo. During that time I had the opportunity to study on exchange at the Hong Kong University of Science and Technology, and work at startups in Toronto and San Francisco as well as a financial services firm in New York.

Really though, I would prefer to spend my time disconnected from technology and the internet – attending film festivals, drinking wine at a beach somewhere warm while reading interesting books, and enjoying the company of good friends. Please talk to me about movies, long-distance running, world events, Georgist socio-economic policy, or literally anything that is not focused on tech.

Testimonials

“Beware of Nitarshan in general… He asks thought-provoking questions that will make you rethink your whole research agenda”

“Quite fun and interesting. Those are his parameters.”

“Nitarshan’s wardrobe is definitely proof of discerning consumption”

Publications

Google Scholar | Semantic Scholar | OpenReview

Self-Supervision for Data Interpretability and Data Efficiency in Reinforcement Learning
MSc Thesis
Nitarshan Rajkumar

Pretraining Representations for Data-Efficient Reinforcement Learning
arXiv Code NeurIPS 2021 ICLR 2021 Workshop (SSL-RL) (Rejected from ICML 2021)
Max Schwarzer, Nitarshan Rajkumar, Michael Noukhovitch, Ankesh Anand, Laurent Charlin, Devon Hjelm, Phil Bachman, Aaron Courville

In Search of Robust Measures of Generalization
arXiv Code NeurIPS 2020
Karolina Dziugaite, Alex Drouin, Brady Neal, Nitarshan Rajkumar, Ethan Caballero, Linbo Wang, Ioannis Mitliagkas, Dan Roy

Blog Posts

Weight Uncertainty in Neural Networks