Tristan Deleu

Mila - Quebec Artificial Intelligence Institute · Université de Montréal

profile_small.jpg

I am a final year Ph.D. candidate at Mila & Université de Montréal, under the supervision of Yoshua Bengio. My research interests include Probabilistic Modeling, Structure Learning, Causal Discovery, Meta-Learning, and Reinforcement Learning. I received the Antidote Scholarship 2019.

Prior to joining Mila, I was a data-scientist at Snips (now acquired by Sonos), working on language understanding for privacy-preserving voice assistants. I also hold a Master in Machine Learning (MVA) from the Ecole Normale Supérieure de Cachan.

Selected publications

  1. UAI
    Discrete Probabilistic Inference as Control in Multi-path Environments
    Tristan Deleu, Padideh Nouri, Nikolay Malkin, Doina Precup, and Yoshua Bengio
    Uncertainty in Artificial Intelligence, 2024
    arXiv
  2. NeurIPS
    Joint Bayesian Inference of Graphical Structure and Parameters with a Single Generative Flow Network
    Tristan Deleu, Mizu Nishikawa-Toomey, Jithendaraa Subramanian, Nikolay Malkin, Laurent Charlin, and Yoshua Bengio
    Conference on Neural Information Processing Systems, 2023
    arXiv
  3. UAI
    Bayesian Structure Learning with Generative Flow Networks
    Tristan Deleu, António Góis, Chris Emezue, Mansi Rankawat, Simon Lacoste-Julien, Stefan Bauer, and Yoshua Bengio
    Uncertainty in Artificial Intelligence, 2022
    arXiv Code
  4. ICLR Spotlight
    Continuous-Time Meta-Learning with Forward Mode Differentiation
    Tristan Deleu, David Kanaa, Leo Feng, Giancarlo Kerg, Yoshua Bengio, Guillaume Lajoie, and Pierre-Luc Bacon
    International Conference on Learning Representations, 2022
    arXiv Code
  5. Preprint Best in Show
    Torchmeta: A Meta-Learning library for PyTorch
    Tristan Deleu, Tobias Würfl, Mandana Samiei, Joseph Paul Cohen, and Yoshua Bengio
    arXiv preprint, 2019
    arXiv Code