Portrait of Yoshua Bengio

Yoshua Bengio

Core Academic Member
Canada CIFAR AI Chair
Full Professor, Université de Montréal, Department of Computer Science and Operations Research Department
Scientific Director, Leadership Team
Observer, Board of Directors, Mila

Biography

*For media requests, please write to medias@mila.quebec.

For more information please contact Julie Mongeau, executive assistant at julie.mongeau@mila.quebec.

Yoshua Bengio is recognized worldwide as a leading expert in AI. He is most known for his pioneering work in deep learning, which earned him the 2018 A.M. Turing Award, “the Nobel Prize of computing,” with Geoffrey Hinton and Yann LeCun.

Bengio is a full professor at Université de Montréal, and the founder and scientific director of Mila – Quebec Artificial Intelligence Institute. He is also a senior fellow at CIFAR and co-directs its Learning in Machines & Brains program, serves as scientific director of IVADO, and holds a Canada CIFAR AI Chair.

In 2019, Bengio was awarded the prestigious Killam Prize and in 2022, he was the most cited computer scientist in the world by h-index. He is a Fellow of the Royal Society of London, Fellow of the Royal Society of Canada, Knight of the Legion of Honor of France and Officer of the Order of Canada. In 2023, he was appointed to the UN’s Scientific Advisory Board for Independent Advice on Breakthroughs in Science and Technology.

Concerned about the social impact of AI, Bengio helped draft the Montréal Declaration for the Responsible Development of Artificial Intelligence and continues to raise awareness about the importance of mitigating the potentially catastrophic risks associated with future AI systems.

Current Students

Professional Master's - Université de Montréal
Co-supervisor :
Professional Master's - Université de Montréal
PhD - Université de Montréal
Postdoctorate - Université de Montréal
Co-supervisor :
Postdoctorate - Université de Montréal
PhD - Université de Montréal
Collaborating researcher - Université Paris-Saclay
Principal supervisor :
Professional Master's - Université de Montréal
Independent visiting researcher - MIT
PhD - École Polytechnique Montréal Fédérale de Lausanne
Research Intern - Université du Québec à Rimouski
Collaborating researcher
Principal supervisor :
PhD - Université de Montréal
Principal supervisor :
Postdoctorate - Université de Montréal
Co-supervisor :
Professional Master's - Université de Montréal
PhD - Université de Montréal
Co-supervisor :
PhD - Université de Montréal
Principal supervisor :
Postdoctorate - Université de Montréal
Co-supervisor :
Master's Research - Université de Montréal
PhD - Université de Montréal
Research Intern - Université de Montréal
Collaborating Alumni
Independent visiting researcher - Université de Montréal
PhD - Université de Montréal
Principal supervisor :
Independent visiting researcher - Université de Montréal
Professional Master's - Université de Montréal
Research Intern - Université de Montréal
PhD - Université de Montréal
PhD - Massachusetts Institute of Technology
PhD - Université de Montréal
PhD - Université de Montréal
Independent visiting researcher - Technical University Munich (TUM)
Independent visiting researcher - Hong Kong University of Science and Technology (HKUST)
DESS - Université de Montréal
Independent visiting researcher - UQAR
Postdoctorate - Université de Montréal
PhD - Université de Montréal
Research Intern - Université de Montréal
Independent visiting researcher - Technical University of Munich
Research Intern - Imperial College London
PhD - Université de Montréal
Co-supervisor :
Postdoctorate - Université de Montréal
PhD - McGill University
Principal supervisor :
Professional Master's - Université de Montréal
Collaborating researcher - Université de Montréal
Research Intern - Université de Montréal
Research Intern - Université de Montréal
PhD - Max-Planck-Institute for Intelligent Systems
PhD - McGill University
Principal supervisor :
Collaborating Alumni - Université de Montréal
Professional Master's - Université de Montréal
PhD - Université de Montréal
Independent visiting researcher - Université de Montréal
Collaborating Alumni - Université de Montréal
Collaborating researcher
Professional Master's - Université de Montréal
Collaborating researcher - Valence
Principal supervisor :
PhD - Université de Montréal
PhD - Université de Montréal
Principal supervisor :
PhD - Université de Montréal
PhD - Université de Montréal
Principal supervisor :
Research Intern - Université de Montréal
Collaborating researcher - Université de Montréal
Independent visiting researcher
Co-supervisor :
Postdoctorate - Université de Montréal
Research Intern - McGill University
Professional Master's - Université de Montréal
Collaborating researcher
Principal supervisor :
Master's Research - Université de Montréal
Co-supervisor :
Master's Research - Université de Montréal
Collaborating researcher - RWTH Aachen University (Rheinisch-Westfälische Technische Hochschule Aachen)
Principal supervisor :
Undergraduate - Université de Montréal
PhD - Université de Montréal
Professional Master's - Université de Montréal
Professional Master's - Université de Montréal
Research Intern - Université de Montréal
PhD - Université de Montréal
Principal supervisor :
Professional Master's - Université de Montréal
Postdoctorate - Université de Montréal

Publications

NEURAL NETWORK-BASED SOLVERS FOR PDES
M. Cameron
Ian G Goodfellow
(1) N (x; θ) = Ll+1 ○ σl ○Ll ○ σl−1 ○ . . . ○ σ1 ○L1. The symbol Lk denotes the k’s affine operator of the form Lk(x) = … (see more)Akx + bk, while σk denotes a nonlinear function called an activation function. The activation functions are chosen by the user. The matrices Ak and shift vectors (or bias vectors) bk are encoded into the argument θ: θ = {Ak, bk} l+1 k=1. The term training neural network means finding {Ak, bk} l+1 k=1 such that N (x; θ) satisfies certain conditions. These conditions are described by the loss function chosen by the user. For example, one might want the neural network to assume certain values fj at certain points xj , j = 1, . . . ,N . These points x are called the training data. In this case, a common choice of the loss function is the least squares error:
Stochastic Generative Flow Networks
Ling Pan
Dinghuai Zhang
Moksh J. Jain
Longbo Huang
Generative Flow Networks (or GFlowNets for short) are a family of probabilistic agents that learn to sample complex combinatorial structures… (see more) through the lens of ``inference as control''. They have shown great potential in generating high-quality and diverse candidates from a given energy landscape. However, existing GFlowNets can be applied only to deterministic environments, and fail in more general tasks with stochastic dynamics, which can limit their applicability. To overcome this challenge, this paper introduces Stochastic GFlowNets, a new algorithm that extends GFlowNets to stochastic environments. By decomposing state transitions into two steps, Stochastic GFlowNets isolate environmental stochasticity and learn a dynamics model to capture it. Extensive experimental results demonstrate that Stochastic GFlowNets offer significant advantages over standard GFlowNets as well as MCMC- and RL-based approaches, on a variety of standard benchmarks with stochastic dynamics.
Stochastic Generative Flow Networks
Ling Pan
Dinghuai Zhang
Moksh J. Jain
Longbo Huang
Supplementary Material for MixupE
Yingtian Zou
Vikas Verma
Sarthak Mittal
Wai Hoh Tang
Hieu Pham
Juho Kannala
Arno Solin
Kenji Kawaguchi
We denote by z = (x,y) the input and output pair where x ∈ X ⊆ R and y ∈ Y ⊆ R . Let fθ(x) ∈ R be the output of the logits (i.e.,… (see more) the last layer before the softmax or sigmoid) of the model parameterized by θ. We use l(θ, z) = h(fθ(x)) − yfθ(x) to denote the loss function. Let g(·) be the activation function. We use x(i) to index i-th element of the vector x and xj to represent j-th variable in a set. The notation list is:
Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learning
Sébastien Lachapelle
Tristan Deleu
Divyat Mahajan
Quentin Bertrand
Although disentangled representations are often said to be beneficial for downstream tasks, current empirical and theoretical understanding … (see more)is limited. In this work, we provide evidence that disentangled representations coupled with sparse base-predictors improve generalization. In the context of multi-task learning, we prove a new identifiability result that provides conditions under which maximally sparse base-predictors yield disentangled representations. Motivated by this theoretical result, we propose a practical approach to learn disentangled representations based on a sparsity-promoting bi-level optimization problem. Finally, we explore a meta-learning version of this algorithm based on group Lasso multiclass SVM base-predictors, for which we derive a tractable dual formulation. It obtains competitive results on standard few-shot classification benchmarks, while each task is using only a fraction of the learned representations.
Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learning
Sébastien Lachapelle
Tristan Deleu
Divyat Mahajan
Quentin Bertrand
Although disentangled representations are often said to be beneficial for downstream tasks, current empirical and theoretical understanding … (see more)is limited. In this work, we provide evidence that disentangled representations coupled with sparse task-specific predictors improve generalization. In the context of multi-task learning, we prove a new identifiability result that provides conditions under which maximally sparse predictors yield disentangled representations. Motivated by this theoretical result, we propose a practical approach to learn disentangled representations based on a sparsity-promoting bi-level optimization problem. Finally, we explore a meta-learning version of this algorithm based on group Lasso multiclass SVM predictors, for which we derive a tractable dual formulation. It obtains competitive results on standard few-shot classification benchmarks, while each task is using only a fraction of the learned representations.
A theory of continuous generative flow networks
Salem Lahlou
Tristan Deleu
Pablo Lemos
Dinghuai Zhang
Alexandra Volokhova
Alex Hernandez-Garcia
Lena Nehale Ezzine
Nikolay Malkin
Generative flow networks (GFlowNets) are amortized variational inference algorithms that are trained to sample from unnormalized target dist… (see more)ributions over compositional objects. A key limitation of GFlowNets until this time has been that they are restricted to discrete spaces. We present a theory for generalized GFlowNets, which encompasses both existing discrete GFlowNets and ones with continuous or hybrid state spaces, and perform experiments with two goals in mind. First, we illustrate critical points of the theory and the importance of various assumptions. Second, we empirically demonstrate how observations about discrete GFlowNets transfer to the continuous case and show strong results compared to non-GFlowNet baselines on several previously studied tasks. This work greatly widens the perspectives for the application of GFlowNets in probabilistic inference and various modeling settings.
A theory of continuous generative flow networks
Salem Lahlou
Tristan Deleu
Pablo Lemos
Dinghuai Zhang
Alexandra Volokhova
Alex Hernandez-Garcia
Lena Nehale Ezzine
Nikolay Malkin
Bayesian learning of Causal Structure and Mechanisms with GFlowNets and Variational Bayes
Mizu Nishikawa-Toomey
Tristan Deleu
Jithendaraa Subramanian
Bayesian causal structure learning aims to learn a posterior distribution over directed acyclic graphs (DAGs), and the mechanisms that defin… (see more)e the relationship between parent and child variables. By taking a Bayesian approach, it is possible to reason about the uncertainty of the causal model. The notion of modelling the uncertainty over models is particularly crucial for causal structure learning since the model could be unidentifiable when given only a finite amount of observational data. In this paper, we introduce a novel method to jointly learn the structure and mechanisms of the causal model using Variational Bayes, which we call Variational Bayes-DAG-GFlowNet (VBG). We extend the method of Bayesian causal structure learning using GFlowNets to learn not only the posterior distribution over the structure, but also the parameters of a linear-Gaussian model. Our results on simulated data suggest that VBG is competitive against several baselines in modelling the posterior over DAGs and mechanisms, while offering several advantages over existing methods, including the guarantee to sample acyclic graphs, and the flexibility to generalize to non-linear causal mechanisms.
Inductive biases for deep learning of higher-level cognition
Anirudh Goyal
Lookback for Learning to Branch
Prateek Gupta
Elias Boutros Khalil
Didier Chételat
M. Pawan Kumar
Towards Scaling Difference Target Propagation by Learning Backprop Targets
Maxence Ernoult
Fabrice Normandin
Abhinav Moudgil
Sean Spinney
The development of biologically-plausible learning algorithms is important for understanding learning in the brain, but most of them fail to… (see more) scale-up to real-world tasks, limiting their potential as explanations for learning by real brains. As such, it is important to explore learning algorithms that come with strong theoretical guarantees and can match the performance of backpropagation (BP) on complex tasks. One such algorithm is Difference Target Propagation (DTP), a biologically-plausible learning algorithm whose close relation with Gauss-Newton (GN) optimization has been recently established. However, the conditions under which this connection rigorously holds preclude layer-wise training of the feedback pathway synaptic weights (which is more biologically plausible). Moreover, good alignment between DTP weight updates and loss gradients is only loosely guaranteed and under very specific conditions for the architecture being trained. In this paper, we propose a novel feedback weight training scheme that ensures both that DTP approximates BP and that layer-wise feedback weight training can be restored without sacrificing any theoretical guarantees. Our theory is corroborated by experimental results and we report the best performance ever achieved by DTP on CIFAR-10 and ImageNet 32