"Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment"
2025 International Conference on Machine Learning (ICML)
Access paper here
I’m a Toronto-based researcher and developer with primary research interests in AI Safety and Interpretability, particularly as they apply to Medicine and Robotics. I'm currently completing my Masters of Science in Applied Computing (MScAC) at the University of Toronto.
In my spare time I try to get outside as much as possible, and watch great and/or terrible movies.
A demo for our Universal Sparse Autoencoder project at ICML 2025! Made in collaboration with Thomas Fel.
We presented our paper "Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment" at ICML 2025 in Vancouver.
2025 International Conference on Machine Learning (ICML)
Access paper here
Proceedings of the 2024 IEEE International Conference on Smart Mobility
Access paper here
@ York University, 2024 - present
- Collaborated with an inter-institutional team to conduct cutting-edge research focused on
interpretability in deep neural networks
- Designed and implemented an innovative training framework
- Developed statistical and visual tools to communicate our findings and reveal model insights
- Contributed as second author to a paper for ICML 2025 (International Conference on Machine Learning)
NSERC Undergraduate Student Research Award
@ York University, 2023
- Developed motion control and obstacle avoidance protocols for an
autonomous wheelchair robot.
- Collaborated with robotics team made up of engineers and developers under Dr.
James Elder.
@ York University, 2022 - Present
- Guided students in Java coding, Object-Oriented Programming
concepts and data structures.
- Helped students debug their code and work through challenges.
@ University of Toronto, Sept 2025 – Dec 2026 (expected)
@ York University, 2020 - 2024, GPA: 3.9
Graduated First Class with Distinction
@ Concordia University, 2017 - 2019, GPA: 3.8
@ York University, 2024
- Exploring Generative Adversarial Networks as a tool for predicting energy usage
on residential power grids. Supervised by Dr. Michael Jenkin.
@ York University, 2023
- Created a tool to improve word choice in text using GPT-2 and
BERT.
@ York University, 2023
- Built a model marketplace with Java Spring, SQL backend, and React
frontend, deployed on AWS.