Get our free extension to see links to code for papers anywhere online!

Chrome logo Add to Chrome

Firefox logo Add to Firefox
Picture for Nicolas Le Roux

Nicolas Le Roux

SIERRA, LIENS

Bridging the Gap Between Adversarial Robustness and Optimization Bias


Feb 17, 2021
Fartash Faghri, Cristina Vasconcelos, David J. Fleet, Fabian Pedregosa, Nicolas Le Roux

Figure 1 for Bridging the Gap Between Adversarial Robustness and Optimization Bias
Figure 2 for Bridging the Gap Between Adversarial Robustness and Optimization Bias
Figure 3 for Bridging the Gap Between Adversarial Robustness and Optimization Bias
Figure 4 for Bridging the Gap Between Adversarial Robustness and Optimization Bias

  Access Paper or Ask Questions

An Effective Anti-Aliasing Approach for Residual Networks


Nov 20, 2020
Cristina Vasconcelos, Hugo Larochelle, Vincent Dumoulin, Nicolas Le Roux, Ross Goroshin

Figure 1 for An Effective Anti-Aliasing Approach for Residual Networks
Figure 2 for An Effective Anti-Aliasing Approach for Residual Networks
Figure 3 for An Effective Anti-Aliasing Approach for Residual Networks
Figure 4 for An Effective Anti-Aliasing Approach for Residual Networks

  Access Paper or Ask Questions

Beyond variance reduction: Understanding the true impact of baselines on policy optimization


Aug 31, 2020
Wesley Chung, Valentin Thomas, Marlos C. Machado, Nicolas Le Roux

Figure 1 for Beyond variance reduction: Understanding the true impact of baselines on policy optimization
Figure 2 for Beyond variance reduction: Understanding the true impact of baselines on policy optimization
Figure 3 for Beyond variance reduction: Understanding the true impact of baselines on policy optimization
Figure 4 for Beyond variance reduction: Understanding the true impact of baselines on policy optimization

  Access Paper or Ask Questions

An operator view of policy gradient methods


Jun 22, 2020
Dibya Ghosh, Marlos C. Machado, Nicolas Le Roux

Figure 1 for An operator view of policy gradient methods
Figure 2 for An operator view of policy gradient methods

  Access Paper or Ask Questions

To Each Optimizer a Norm, To Each Norm its Generalization


Jun 11, 2020
Sharan Vaswani, Reza Babanezhad, Jose Gallego, Aaron Mishkin, Simon Lacoste-Julien, Nicolas Le Roux

Figure 1 for To Each Optimizer a Norm, To Each Norm its Generalization
Figure 2 for To Each Optimizer a Norm, To Each Norm its Generalization
Figure 3 for To Each Optimizer a Norm, To Each Norm its Generalization
Figure 4 for To Each Optimizer a Norm, To Each Norm its Generalization

  Access Paper or Ask Questions

The Geometry of Sign Gradient Descent


Feb 19, 2020
Lukas Balles, Fabian Pedregosa, Nicolas Le Roux

Figure 1 for The Geometry of Sign Gradient Descent
Figure 2 for The Geometry of Sign Gradient Descent
Figure 3 for The Geometry of Sign Gradient Descent
Figure 4 for The Geometry of Sign Gradient Descent

  Access Paper or Ask Questions

Information matrices and generalization


Jun 18, 2019
Valentin Thomas, Fabian Pedregosa, Bart van Merriënboer, Pierre-Antoine Mangazol, Yoshua Bengio, Nicolas Le Roux

Figure 1 for Information matrices and generalization
Figure 2 for Information matrices and generalization
Figure 3 for Information matrices and generalization
Figure 4 for Information matrices and generalization

  Access Paper or Ask Questions

Reducing the variance in online optimization by transporting past gradients


Jun 18, 2019
Sébastien M. R. Arnold, Pierre-Antoine Manzagol, Reza Babanezhad, Ioannis Mitliagkas, Nicolas Le Roux

Figure 1 for Reducing the variance in online optimization by transporting past gradients
Figure 2 for Reducing the variance in online optimization by transporting past gradients
Figure 3 for Reducing the variance in online optimization by transporting past gradients
Figure 4 for Reducing the variance in online optimization by transporting past gradients
* Open-source implementation available at: https://github.com/seba-1511/igt.pth 
  Access Paper or Ask Questions

Anytime Tail Averaging


Feb 19, 2019
Nicolas Le Roux

Figure 1 for Anytime Tail Averaging
Figure 2 for Anytime Tail Averaging
Figure 3 for Anytime Tail Averaging
* Added a specific section on the case of multiple accumulators when k_t is a constant 
  Access Paper or Ask Questions

The Value Function Polytope in Reinforcement Learning


Feb 15, 2019
Robert Dadashi, Adrien Ali Taïga, Nicolas Le Roux, Dale Schuurmans, Marc G. Bellemare

Figure 1 for The Value Function Polytope in Reinforcement Learning
Figure 2 for The Value Function Polytope in Reinforcement Learning
Figure 3 for The Value Function Polytope in Reinforcement Learning
Figure 4 for The Value Function Polytope in Reinforcement Learning

  Access Paper or Ask Questions

Distributional reinforcement learning with linear function approximation


Feb 08, 2019
Marc G. Bellemare, Nicolas Le Roux, Pablo Samuel Castro, Subhodeep Moitra

Figure 1 for Distributional reinforcement learning with linear function approximation
Figure 2 for Distributional reinforcement learning with linear function approximation
* Proceedings of AISTATS 2019 
 * To appear 
  Access Paper or Ask Questions

Negative eigenvalues of the Hessian in deep neural networks


Feb 06, 2019
Guillaume Alain, Nicolas Le Roux, Pierre-Antoine Manzagol

Figure 1 for Negative eigenvalues of the Hessian in deep neural networks
Figure 2 for Negative eigenvalues of the Hessian in deep neural networks
Figure 3 for Negative eigenvalues of the Hessian in deep neural networks
Figure 4 for Negative eigenvalues of the Hessian in deep neural networks

  Access Paper or Ask Questions

A Geometric Perspective on Optimal Representations for Reinforcement Learning


Jan 31, 2019
Marc G. Bellemare, Will Dabney, Robert Dadashi, Adrien Ali Taiga, Pablo Samuel Castro, Nicolas Le Roux, Dale Schuurmans, Tor Lattimore, Clare Lyle

Figure 1 for A Geometric Perspective on Optimal Representations for Reinforcement Learning
Figure 2 for A Geometric Perspective on Optimal Representations for Reinforcement Learning
Figure 3 for A Geometric Perspective on Optimal Representations for Reinforcement Learning
Figure 4 for A Geometric Perspective on Optimal Representations for Reinforcement Learning

  Access Paper or Ask Questions

Understanding the impact of entropy on policy optimization


Nov 29, 2018
Zafarali Ahmed, Nicolas Le Roux, Mohammad Norouzi, Dale Schuurmans

Figure 1 for Understanding the impact of entropy on policy optimization
Figure 2 for Understanding the impact of entropy on policy optimization
Figure 3 for Understanding the impact of entropy on policy optimization
Figure 4 for Understanding the impact of entropy on policy optimization

  Access Paper or Ask Questions

Tracking the gradients using the Hessian: A new look at variance reducing stochastic methods


Mar 31, 2018
Robert M. Gower, Nicolas Le Roux, Francis Bach

Figure 1 for Tracking the gradients using the Hessian: A new look at variance reducing stochastic methods
Figure 2 for Tracking the gradients using the Hessian: A new look at variance reducing stochastic methods
Figure 3 for Tracking the gradients using the Hessian: A new look at variance reducing stochastic methods
* 17 pages, 2 figures, 1 table 
  Access Paper or Ask Questions

Distributed SAGA: Maintaining linear convergence rate with limited communication


May 29, 2017
Clément Calauzènes, Nicolas Le Roux

Figure 1 for Distributed SAGA: Maintaining linear convergence rate with limited communication
Figure 2 for Distributed SAGA: Maintaining linear convergence rate with limited communication
Figure 3 for Distributed SAGA: Maintaining linear convergence rate with limited communication

  Access Paper or Ask Questions

A comparative study of counterfactual estimators


May 02, 2017
Thomas Nedelec, Nicolas Le Roux, Vianney Perchet

Figure 1 for A comparative study of counterfactual estimators
Figure 2 for A comparative study of counterfactual estimators
Figure 3 for A comparative study of counterfactual estimators
Figure 4 for A comparative study of counterfactual estimators

  Access Paper or Ask Questions

Tighter bounds lead to improved classifiers


Dec 28, 2016
Nicolas Le Roux

Figure 1 for Tighter bounds lead to improved classifiers
Figure 2 for Tighter bounds lead to improved classifiers
Figure 3 for Tighter bounds lead to improved classifiers
Figure 4 for Tighter bounds lead to improved classifiers

  Access Paper or Ask Questions

Efficient iterative policy optimization


Dec 28, 2016
Nicolas Le Roux

Figure 1 for Efficient iterative policy optimization
Figure 2 for Efficient iterative policy optimization
Figure 3 for Efficient iterative policy optimization
* 12 pages 
  Access Paper or Ask Questions

Minimizing Finite Sums with the Stochastic Average Gradient


May 11, 2016
Mark Schmidt, Nicolas Le Roux, Francis Bach

Figure 1 for Minimizing Finite Sums with the Stochastic Average Gradient
Figure 2 for Minimizing Finite Sums with the Stochastic Average Gradient
Figure 3 for Minimizing Finite Sums with the Stochastic Average Gradient
Figure 4 for Minimizing Finite Sums with the Stochastic Average Gradient
* Revision from January 2015 submission. Major changes: updated literature follow and discussion of subsequent work, additional Lemma showing the validity of one of the formulas, somewhat simplified presentation of Lyapunov bound, included code needed for checking proofs rather than the polynomials generated by the code, added error regions to the numerical experiments 
  Access Paper or Ask Questions

A Stochastic Gradient Method with an Exponential Convergence Rate for Finite Training Sets


Mar 11, 2013
Nicolas Le Roux, Mark Schmidt, Francis Bach

Figure 1 for A Stochastic Gradient Method with an Exponential Convergence Rate for Finite Training Sets
* The notable changes over the current version: - worked example of convergence rates showing SAG can be faster than first-order methods - pointing out that the storage cost is O(n) for linear models - the more-stable line-search - comparison to additional optimal SG methods - comparison to rates of coordinate descent methods in quadratic case 
  Access Paper or Ask Questions

Local Component Analysis


Dec 10, 2012
Nicolas Le Roux, Francis Bach

Figure 1 for Local Component Analysis
Figure 2 for Local Component Analysis
Figure 3 for Local Component Analysis
Figure 4 for Local Component Analysis

  Access Paper or Ask Questions

Convergence Rates of Inexact Proximal-Gradient Methods for Convex Optimization


Dec 01, 2011
Mark Schmidt, Nicolas Le Roux, Francis Bach

Figure 1 for Convergence Rates of Inexact Proximal-Gradient Methods for Convex Optimization
Figure 2 for Convergence Rates of Inexact Proximal-Gradient Methods for Convex Optimization
* Neural Information Processing Systems (2011) 
  Access Paper or Ask Questions

Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs


Jul 19, 2011
Nicolas Heess, Nicolas Le Roux, John Winn

Figure 1 for Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs
Figure 2 for Weakly Supervised Learning of Foreground-Background Segmentation using Masked RBMs
* International Conference on Artificial Neural Networks (2011) 
  Access Paper or Ask Questions