News

October 2026I am recruiting PhD and MSc students for Fall 2027. See how to apply.
May 2026Hiroki Naganuma has completed his PhD and joins NVIDIA. Congratulations Dr. Naganuma!
May 2026Two papers accepted at ICML 2026, on unlearning with public data and on adaptive batch sizes for sign and spectral descent.
January 2026Three papers accepted at ICLR 2026, including Beyond Multi-Token Prediction: Pretraining LLMs with Future Summaries.
September 2025Understanding Adam Requires Better Rotation Dependent Assumptions accepted at NeurIPS 2025.
May 2025Compositional Risk Minimization accepted at ICML 2025.
January 2025Charles Guille-Escuret has completed his PhD and joins the MBZUAI Institute of Foundation Models. Congratulations Dr. Guille-Escuret!
June 2024One more strong PhD graduation, this time it is Adam Ibrahim! He continues his work as a research scientist at H, Paris. Congratulations Dr. Ibrahim!
May 2024No Wrong Turns: The Simple Geometry Of Neural Networks Optimization Paths paper accepted at ICML 2024!
May 2024For the first time in a long while, I'm attending a big ML conference. I will be at ICLR 2024 in Vienna.
April 2024My second PhD student to graduate, Reyhane Askari Hemmat, has successfully defended her thesis. She continues her work as a research scientist at Meta, Montreal. Congratulations Dr. Askari!
January 2024Empirical Analysis of Model Selection for Heterogeneous Causal Effect Estimation paper accepted at ICLR 2024 for an spotlight presentation.
January 2024LEAD: Min-Max Optimization from a Physical Perspective invited to be presented at ICLR 2024
December 2023For the winter semester of 2024, I am on sabbatical at Archimedes in Athens, Greece.
September 2023Additive Decoders for Latent Variables Identification and Cartesian-Product Extrapolation paper accepted at NeurIPS 2023 for an oral presentation
September 2023CADet: Fully Self-Supervised Out-Of-Distribution Detection With Contrastive Learning paper accepted at NeurIPS 2023
September 2023Welcome to new PhD student, Zichu Liu.
September 2023Welcome to new intern, Ange-Clement Akazan who's visitting us from the African Institute for Mathematical Research and will study generative models
April 2023Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learning paper accepted at ICML 2023.
January 2023My first PhD student to graduate, Alexia Jolicoeur-Martineau, has successfully defended her thesis. She continues her work as a research scientist at SAIL Montreal. Congratulations Dr. Jolicoeur-Martineau!!
January 2023Performative Prediction with Neural Networks paper accepted at AISTASTS 2023.
January 2023Neural Networks Efficiently Learn Low-Dimensional Representations with SGD paper accepted at ICLR 2023 with honorable mention.
September 2022I'm excited to announce that I've started part-time work as a staff research scientist at Google Brain Montreal! Looking forward to deepening my connection to larger scale problems in industry.
September 2022Gradient Descent Is Optimal Under Lower Restricted Secant Inequality And Upper Error Bound paper accepted at NeurIPS 2022.
September 2022Kartik Ahuja has finished his postdoc and started working as a research scientist at Meta research (FAIR) Paris!
July 2022Graduating PhD candidate Alexia Jolicoeur-Martineau started work at Samsung SAIT AI Lab in Montreal!
June 2022As of June 1st, 2022 I was promoted to the post of associate professor with tenure. My sincere thanks to colleagues and my group for all their hard work.
October 2021Honored to be invited to the semester of Learning and Games at Simons. I will be visitting Berkeley in January-March of 2022.
September 2021Welcome to new MSc students, Mehrnaz Mofakhami and Divyat Mahajan.
September 2021Two papers accepted at NeurIPS 2021! [one, two]
April 2021Postdoc in our group, Nicolas Loizou accepted a tenure track position at Johns Hopkins University!
March 2021Postdoc in our group, Manuela Girotti accepted a tenure track position at St. Mary's University! In the meantime shewill spend time at the prestigious mathematical institute, MSRI, in Berkeley, CA.
January 2021Best student paper award for Charles, Baptiste and Manuela's paper at OPT2020! For their work on the fundamentals of condition numbers. Their paper was accepted for publication at AISTATS 2021.
January 2021Alexia and Rémi's paper, in collaboration with MSR Montréal has been accepted at ICLR 2021! Preprint available here.
September 2020Paper on evaluating generalization measures accepted at NeurIPS'20.
September 2020Welcome to new PhD students, Ryan D'Orazio and Hiroki Naganuma.
August 2020Kartik Ahuja awarded the IVADO postdoctoral scholarship. Excited to have him join us in January 2021!
April 2020Two papers on differentiable games (one, two) accepted at ICML'20.
January 2020Two papers on efficient methods and tight bounds for differentiable games (one, two) accepted at AISTATS'20.
December 2019Nicolas Loizou was awarded the IVADO postdoctoral scholarship at the prestigious Fellow tier.
December 2019Brady Neal graduates with an MSc. He will continue on his PhD with us.
November 2019Excited to be coorganizing the 2nd iteration of the Smooth Games Optimization and Machine Learning Workshop at NeurIPS'19.
November 2019Reducing the variance in online optimization by transporting past gradients selected for spotlight oral presentation at NeurIPS'19.
October 2019Multiple submissions to AISTATS. Preprints on the way...
June 2019At ICML with 3 papers in main conference, 2 in Deep Learning Phenomena workshop.
May 2019State-Reification Networks selected for oral presentation at ICML'19.
April 2019Excited to be listed among the most prolific authors of accepted ICML 2019 papers.
April 2019I received the NSERC Discovery grant!
April 2019In Japan for AISTATS.
January 2019h-detach paper accepted at ICLR.
December 2018Was nominated in the first cohort of Canada CIFAR AI chairs!!
December 2018Co-organizing Smooth Games Optimization in ML workshop at NeurIPS.
December 2018Negative momentum for improved game dynamics. Paper accepted at AISTATS 2019.
December 2018Full version of YellowFin manscript accepted at SysML
September 2018Excited to be teaching Machine Learning to a class of 180 graduate students at UdeM.
February 2018YellowFin selected for oral presentation at SysML'18.
January 2018Teaching new class! IFT 6085: Theoretical principles for deep learning
December 2017Accelerated power iteration via momentum, paper accepted at AISTATS 2018.
November 2017Talk at Google Brain, Montréal
September 2017Thrilled to be starting work at the University of Montreal and the Mila as an assistant professor!
August 2017Visiting my alma mater, UT Austin.
August 2017At Sydney for ICML, presenting work on YellowFin, custom scans for Gibbs sampling, and deep learning for 3D point cloud representation and generation.
July 2017New preprint! Representation Learning and Adversarial Generation of 3D Point Clouds [arxiv].
July 2017New preprint! Accelerated stochastic power iteration [arxiv].
June 2017New preprint! An automatic tuner for they hyperparameters of momentum SGD [arxiv].
May 2017Custom scan sequence paper accepted for presentation at ICML 2017!
April 2017Invited talk at Workshop on Advances in Computing Architectures, Stanford SystemX
March 2017New preprint! Custom scan sequences for super fast Gibbs sampling.
February 2017Invited to talk at ITA in San Diego.
February 2017Spoke at the AAAI 2017 Workshop on Distributed Machine Learning.
January 2017Visiting Microsoft Research, Cambridge
December 2016At NIPS, presenting our Gibbs sampling paper dispelling some common beliefs regarding scan orders.
November 2016Visiting Microsoft Research New England
November 2016Invited talk at SystemX Stanford Alliance Fall Conference
November 2016Full version of asynchrony paper.
September 2016Talk at Allerton
August 2016I had the pleasure to give a talk MIT Lincoln Labs.
August 2016Gave an asynchronous optimization talk at Google.
August 2016Blog post on our momentum work.
July 2016Invited to talk at NVIDIA.
June 2016Poster at non-convex optimization ICML workshop.
June 2016Poster at OptML 2016 workshop.
2016In a recent note, we show that asynchrony in SGD introduces momentum. In the companion systems paper, we use this theory to train deep networks faster.
2016Does periodic model averaging always help? Recent results.
2015Excited to start Postdoc at Stanford University. Will be working with Lester Mackey and Chris Ré.
2015Successfully defended my PhD thesis!
2015SILO seminar talk at the Wisconsin Institute of Discovery.
2015Densest k-Subgraph work picked up by NVIDIA!
2015Our latest work has been accepted for presentation at VLDB 2015!