Publications (20)
Synergies between Disentanglement and Sparsity: Generalization and Identifiability in Multi-Task Learning
Sébastien Lachapelle, Tristan Deleu, Divyat Mahajan +4
Although disentangled representations are often said to be beneficial for downstream tasks, current empirical and theoretical understanding is limited. In this work, we provide evi…
On the Closed-Form of Flow Matching: Generalization Does Not Arise from Target Stochasticity
Quentin Bertrand, Anne Gagneux, Mathurin Massias +1
Modern deep generative models can now produce high-quality synthetic samples that are often indistinguishable from real training data. A growing body of research aims to understand…
Handling correlated and repeated measurements with the smoothed multivariate square-root Lasso
Quentin Bertrand, Mathurin Massias, Alexandre Gramfort +1
Sparsity promoting norms are frequently used in high dimensional regression. A limitation of such Lasso-type estimators is that the optimal regularization parameter depends on the…
Anderson acceleration of coordinate descent
Quentin Bertrand, Mathurin Massias
Acceleration of first order methods is mainly obtained via inertial techniques à la Nesterov, or via nonlinear extrapolation. The latter has known a recent surge of interest, with…
Anytime Exact Belief Propagation
Gabriel Azevedo Ferreira, Quentin Bertrand, Charles Maussion +1
Statistical Relational Models and, more recently, Probabilistic Programming, have been making strides towards an integration of logic and probabilistic reasoning. A natural expecta…
On the Limitations of Elo: Real-World Games, are Transitive, not Additive
Quentin Bertrand, Wojciech Marian Czarnecki, Gauthier Gidel
Real-world competitive games, such as chess, go, or StarCraft II, rely on Elo models to measure the strength of their players. Since these games are not fully transitive, using Elo…
Dimension improvement in Dhar's refutation of the Eden conjecture
Quentin Bertrand, Jules Pertinand
We consider the Eden model on the d-dimensional hypercubical unoriented lattice , for large d. Initially, every lattice point is healthy, except the origin which is infected. Then,…
Multipacting mitigation by atomic layer deposition: the case study of Titanium Nitride
Yasmine Kalboussi, Sarah Dadouch, Baptiste Delatte +15
This study investigates the use of Atomic Layer deposition (ALD) to mitigate multipacting phenomena inside superconducting radio frequency (SRF) cavities used in particle accelerat…
Beyond L1: Faster and Better Sparse Models with skglm
Quentin Bertrand, Quentin Klopfenstein, Pierre-Antoine Bannier +2
We propose a new fast algorithm to estimate any sparse generalized linear model with convex or non-convex separable penalties. Our algorithm is able to solve problems with millions…
The Control System of the Elliptical Cavity and Cryomodule Test Stand Demonstrator for ESS
Alexis Gaget, Tom Joannem, Adelino Gomes +3
CEA IRFU Saclay is taking part of ESS (European Spallation Source) construction through several packages and, especially in the last three years on the Elliptical Cavity and Cryomo…
Implicit differentiation of Lasso-type models for hyperparameter optimization
Quentin Bertrand, Quentin Klopfenstein, Mathieu Blondel +3
Setting regularization parameters for Lasso-type estimators is notoriously difficult, though crucial in practice. The most popular hyperparameter optimization approach is grid-sear…
Self-Consuming Generative Models with Curated Data Provably Optimize Human Preferences
Damien Ferbach, Quentin Bertrand, Avishek Joey Bose +1
The rapid progress in generative models has resulted in impressive leaps in generation quality, blurring the lines between synthetic and real data. Web-scale datasets are now prone…
The Curse of Unrolling: Rate of Differentiating Through Optimization
Damien Scieur, Quentin Bertrand, Gauthier Gidel +1
Computing the Jacobian of the solution of an optimization problem is a central problem in machine learning, with applications in hyperparameter optimization, meta-learning, optimiz…
On the Stability of Iterative Retraining of Generative Models on their own Data
Quentin Bertrand, Avishek Joey Bose, Alexandre Duplessis +2
Deep generative models have made tremendous progress in modeling complex data, often exhibiting generation quality that surpasses a typical human's ability to discern the authentic…
Omega: Optimistic EMA Gradients
Juan Ramirez, Rohan Sukumaran, Quentin Bertrand +1
Stochastic min-max optimization has gained interest in the machine learning community with the advancements in GANs and adversarial training. Although game optimization is fairly w…
Electromagnetic neural source imaging under sparsity constraints with SURE-based hyperparameter tuning
Pierre-Antoine Bannier, Quentin Bertrand, Joseph Salmon +1
Estimators based on non-convex sparsity-promoting penalties were shown to yield state-of-the-art solutions to the magneto-/electroencephalography (M/EEG) brain source localization…
Model identification and local linear convergence of coordinate descent
Quentin Klopfenstein, Quentin Bertrand, Alexandre Gramfort +2
For composite nonsmooth optimization problems, Forward-Backward algorithm achieves model identification (e.g. support identification for the Lasso) after a finite number of iterati…
Self-Play Q-learners Can Provably Collude in the Iterated Prisoner's Dilemma
Quentin Bertrand, Juan Duque, Emilio Calvano +1
A growing body of computational studies shows that simple machine learning agents converge to cooperative behaviors in social dilemmas, such as collusive price-setting in oligopoly…
Implicit differentiation for fast hyperparameter selection in non-smooth convex learning
Quentin Bertrand, Quentin Klopfenstein, Mathurin Massias +4
Finding the optimal hyperparameters of a model can be cast as a bilevel optimization problem, typically solved using zero-order techniques. In this work we study first-order method…
Support recovery and sup-norm convergence rates for sparse pivotal estimation
Mathurin Massias, Quentin Bertrand, Alexandre Gramfort +1
In high dimensional sparse regression, pivotal estimators are estimators for which the optimal regularization parameter is independent of the noise level. The canonical pivotal est…