Spencer Szabados

My last name is pronounced 'za-ba-dosh'.

prof_pic.jpg

Honolulu, Hawaii (ICML2023)

I am an incoming PhD student at UBC (2025.09 - Present) within the PLAI lab, and graduate intern at Inverted AI. Previously, I was working full time (2024.08 - 2025.08) at hum.ai as a Machine Learning Engineer using diffusion models for temporal prediction of earth remote sensing data.

My research interests are centered around generative modeling with an emphasis on diffusion models applied to robotics and computer vision; I’m also interested in a variety of topics within optimization, probabilistic modeling, and rendering methods.

I hold a Master’s of Mathematics in Computer Science from the University of Waterloo, where I was supervised by Yao-liang Yu working on Machine Learning Generative Models. I am an alumnus of the Vector Institute. Before this, I completed an undergraduate degree at the University of Manitoba in Applied Mathematics with Computer Science; where I was a member of the GADA Lab studying notions of Data Depth, supervised by Stephane Durocher.

Outside of research I enjoy cycling, camping, and some amateur photography. I am also an avid reader, biased towards hard sci-fi that provide interesting social political commentary.

Latest posts

Selected publications

  1. LuSY2024_spigm.png
    Diffusion Models with Group Equivariance
    Haoye Lu , Spencer Szabados, and Yaoliang Yu
    In ICML Workshop SPIGM , 2024
  2. LuLJSSY2023.png
    CM-GAN: Stabilizing GAN Training with Consistency Models
    Haoye Lu , Yiwei Lu , Dihong Jiang , and 3 more authors
    In ICML 2023 Workshop on Structured Probabilistic Inference & Generative Modeling , 2023
  3. DurocherS2022.png
    Curve Stabbing Depth: Data Depth for Plane Curves
    Stephane Durocher , and Spencer Szabados
    In 34th Canadian Conference on Computational Geometry (CCCG 2022) , 2022