Guillaume Lajoie

Membre Académique Principal
Guillaume Lajoie
Professeur adjoint, Université de Montréal
Guillaume Lajoie

Guillaume Lajoie est professeur adjoint au Département de mathématiques et de statistique (DMS) de l’Université de Montréal et membre académique principal de Mila.  Il est également titulaire d’une position de chercheur boursier du FRQS. Il était auparavant boursier postdoctoral au Max Planck Institute for Dynamics and Self-Organization et à l’Université de Washington Institute for Neuroengineering et il a obtenu son doctorat au Département de mathématiques appliquées de l’Université de Washington, à Seattle.

Sa recherche est centrée à l’intersection de l’IA et des neurosciences où il s’intéresse à des questions reliées à la dynamiques et aux computations de réseaux neuronaux, avec certaines applications à la neuroingénérie. Ses travaux récents comprennent le développement de biais inductifs pour un meilleure propagation d’information dans les réseaux récurrents, ainsi que le développement d’algorithmes pour optimiser les interfaces cerveau-machine bidirectionnelles.

Publications

2020-10

LEAD: Least-Action Dynamics for Min-Max Optimization
Reyhane Askari Hemmat, Amartya Mitra, Guillaume Lajoie and Ioannis Mitliagkas
arXiv preprint arXiv:2010.13846
(2020-10-26)
arxiv.orgPDF

2020-09

Predictive learning as a network mechanism for extracting low-dimensional latent space representations
Stefano Recanatesi, Matthew Farrell, Guillaume Lajoie, Sophie Deneve, Mattia Rigotti and Eric Shea-Brown
bioRxiv
(2020-09-17)
www.biorxiv.orgPDF

2020-08

Implicit Regularization via Neural Feature Alignment.
Aristide Baratin, Thomas George, César Laurent, R Devon Hjelm, Guillaume Lajoie, Pascal Vincent and Simon Lacoste-Julien
arXiv: Learning
(2020-08-03)
arxiv.orgPDF
Implicit Regularization in Deep Learning: A View from Function Space
Aristide Baratin, Thomas George, César Laurent, R Devon Hjelm, Guillaume Lajoie, Pascal Vincent and Simon Lacoste-Julien
arXiv preprint arXiv:2008.00938
(2020-08-03)
www.microsoft.comPDF

2020-07

Learning to Combine Top-Down and Bottom-Up Signals in Recurrent Neural Networks with Attention over Modules
Sarthak Mittal, Alex Lamb, Anirudh Goyal, Vikram Voleti, Murray Shanahan, Guillaume Lajoie, Michael Mozer and Yoshua Bengio

2020-06

On Lyapunov Exponents for RNNs: Understanding Information Propagation Using Dynamical Systems Tools
Ryan Vogt, Maximilian Puelma Touzel, Eli Shlizerman and Guillaume Lajoie
arXiv: Learning
(2020-06-25)
ui.adsabs.harvard.eduPDF
Advantages of biologically-inspired adaptive neural activation in RNNs during learning.
Victor Geadah, Giancarlo Kerg, Stefan Horoi, Guy Wolf and Guillaume Lajoie
arXiv preprint arXiv:2006.12253
(2020-06-22)
dblp.uni-trier.dePDF
Untangling tradeoffs between recurrence and self-attention in neural networks
Giancarlo Kerg, Bhargav Kanuparthi, Anirudh Goyal, Kyle Goyette, Yoshua Bengio and Guillaume Lajoie
arXiv preprint arXiv:2006.09471
(2020-06-16)
dblp.uni-trier.dePDF

2020-05

Low-Dimensional Dynamics of Encoding and Learning in Recurrent Neural Networks
Stefan Horoi, Victor Geadah, Guy Wolf and Guillaume Lajoie
Canadian Conference on AI
(2020-05-13)
link.springer.com

2020-04

Hierarchical Bayesian Optimization of Spatiotemporal Neurostimulations for Targeted Motor Outputs
Samuel Laferriere, Marco Bonizzato, Sandrine L. Cote, Numa Dancause and Guillaume Lajoie
IEEE Transactions on Neural Systems and Rehabilitation Engineering
(2020-04-13)
ieeexplore.ieee.org

2020-01

Internal representation dynamics and geometry in recurrent neural networks.
Stefan Horoi, Guillaume Lajoie and Guy Wolf
arXiv preprint arXiv:2001.03255
(2020-01-09)
dblp.uni-trier.dePDF

2019-12

Non-normal Recurrent Neural Network (nnRNN): learning long time dependencies while improving expressivity with transient dynamics
Giancarlo Kerg, Kyle Goyette, Maximilian Puelma Touzel, Gauthier Gidel, Eugene Vorontsov, Yoshua Bengio and Guillaume Lajoie

2019-09

Modelling Working Memory using Deep Recurrent Reinforcement Learning
Pravish Sainath, Pierre Bellec and Guillaume Lajoie
(venue unknown)
(2019-09-11)
openreview.netPDF
Recurrent neural networks learn robust representations by dynamically balancing compression and expansion
Matthew Farrell, Stefano Recanatesi, Guillaume Lajoie and Eric Shea-Brown
bioRxiv
(2019-09-11)
openreview.netPDF

2019-07

Predictive learning extracts latent space representations from sensory observations
Stefano Recanatesi, Matthew Farrell, Guillaume Lajoie, Sophie Deneve, Mattia Rigotti and Eric Shea-Brown
bioRxiv
(2019-07-13)
www.biorxiv.orgPDF

2019-06

Learning to evoke complex motor outputs with spatiotemporal neurostimulation using a hierarchical and adaptive optimization algorithm
Samuel Laferriere, Marco Bonizzato, Numa Dancause and Guillaume Lajoie
bioRxiv
(2019-06-11)
www.biorxiv.orgPDF
Dimensionality compression and expansion in Deep Neural Networks
Stefano Recanatesi, Matthew Farrell, Madhu Advani, Timothy Moore, Guillaume Lajoie and Eric Shea-Brown
arXiv preprint arXiv:1906.00443
(2019-06-02)
ui.adsabs.harvard.eduPDF

2019-01

Cortical network mechanisms of anodal and cathodal transcranial direct current stimulation in awake primates
Andrew R. Bogaard, Guillaume Lajoie, Hayley Boyd, Andrew Morse, Stavros Zanos and Eberhard E. Fetz
bioRxiv
(2019-01-09)
www.biorxiv.orgPDF
Learning to evoke complex motor outputs with spatiotemporal neurostimulation using a hierarchical and adaptive optimization algorithm.
Samuel Laferriere, Guillaume Lajoie, Numa Dancause and Marco Bonizzato
2019 Conference on Cognitive Computational Neuroscience
(2019-01-01)
dx.doi.org

2018-11

Signatures and mechanisms of low-dimensional neural predictive manifolds
Stefano Recanatesi, Matthew Farrell, Guillaume Lajoie, Sophie Deneve, Mattia Rigotti and Eric Shea-Brown
bioRxiv
(2018-11-17)
www.biorxiv.orgPDF

Publications collected and formatted using Paperoni

array(1) { ["wp-wpml_current_language"]=> string(2) "fr" }