Vaden Masrani


Department of Computer Science

University of British Columbia

X725-2366 Main Mall

I’m a Ph.D student working in machine learning and probabilistic programming at the University of British Columbia. I am a member of the PLAI lab under the supervision of Frank Wood. My CV is here.

Before this I was supervised by Giuseppe Carenini in the Laboratory for Computational Intelligence where I completed my M.Sc. in artificial intelligence. I have also had the good fortune of working with Siamak Ravanbakhsh, Emtiyaz Khan, and Colin Gay in deep learning, approximate bayesian inference, and particle physics, respectively.

(And occasional blogger and podcaster on the side).


Dec 12, 2020 We received best paper award at the NeurIPS Workshop on Deep Learning through Information Geometry! Check out Annealed Importance Sampling with q-Paths where we generalize the 1-d geometric path commonly used in Annealed Importance Sampling into a 2-d plane. Poster, code.
Dec 10, 2020 Excited about this one! Gaussian Process Bandit Optimization of the Thermodynamic Variational Objective where we address the problem of scheduling in the TVO using Gaussian Processes and Bandits!
Jul 6, 2020 Excited to announce our new ICML paper All in the Exponential Family: Bregman Duality in Thermodynamic Variational Inference with the brilliant Rob Brekelmans (@brekelmaniac)!
Mar 30, 2020 New preprint! Planning as Inference in Epidemiological Models
Feb 24, 2020 Our paper Improved Few-Shot Visual Classification accepted at CVPR 2020 - great work Peyman!