avery hw ryoo

I am a PhD student at Mila - Quebec AI Institute and the Université de Montréal, where I study Computer Science with a specialization in Artificial Intelligence. I have the fortune to be advised by Dr. Matthew Perich and Dr. Guillaume Lajoie. Check out their respective labs here and here!

I am primarily interested in deep generative models – especially their applications in medicine and science. I aim to use ideas from deep learning, compositionality, and dynamical systems theory to design generative frameworks that are faster and more sample-efficient — a crucial step in mitigating the widening resource disparity in an era of increasingly large models.

In another life, I studied biomedical engineering at the University of Waterloo, during which I completed several internships in data science, computer vision research, and brain-computer interfaces.

research interestsnon-research interests
- deep generative models
- AI for medicine/science
- compositionality/modularity
- dynamical systems
- neuroprosthetics
- light-roast coffee
- watching sports (NBA, soccer, F1, NFL)
- bouldering
- ginger tea
- trivia

news

  • 2025-09-18: our submission, "Towards a generalizable, unified framework for decoding from multimodal neural activity", has been accepted to the BrainBodyFM Workshop at NeurIPS 2025!
  • 2025-09-18: our submission, "Generalizable, real-time neural decoding with hybrid state-space models", has been accepted to NeurIPS 2025! see you in San Diego 🏖️
  • 2025-09-01: attended the PAISS 2025 summer school in Grenoble 🇫🇷
  • 2025-08-25: officially started my PhD!
  • 2025-04-21: our submission, "Generalizable, real-time neural decoding with hybrid state-space models", has been accepted to CCN 2025!
  • 2025-04-11: presented our work on "A generalizable approach to real-time neural decoding" at the Mila Neural-AI Reading Group
  • 2025-02-26: presented our work on "A generalizable approach to real-time neural decoding" at the UNIQUE Fellows Get-Together
  • 2025-01-22: our paper "Expressivity of Neural Networks with Random Weights and Learned Biases" has been accepted to ICLR 2025!
  • 2025-01-20: received the UNIQUE Excellence Scholarship at the master's level. Thank you to UNIQUE for this honour!
  • 2025-01-20: served as a reviewer for CogSURF 2025
  • 2024-12-28: new site!
  • 2024-12-24: our submission, "Towards generalizable, real-time decoders for brain-computer interfaces" has been accepted to Cosyne 2025. See you in Montréal!
  • 2024-12-11: our workshop proposal on foundation models for neuroscience has been accepted to Cosyne 2025 -- we hope to see you there!