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 Florent Krzakala

Classifying high-dimensional Gaussian mixtures: Where kernel methods fail and neural networks succeed


Feb 23, 2021
Maria Refinetti, Sebastian Goldt, Florent Krzakala, Lenka Zdeborová

* The accompanying code for this paper is available at https://github.com/mariaref/rfvs2lnn_GMM_online 

  Access Paper or Ask Questions

Capturing the learning curves of generic features maps for realistic data sets with a teacher-student model


Feb 16, 2021
Bruno Loureiro, Cédric Gerbelot, Hugo Cui, Sebastian Goldt, Florent Krzakala, Marc Mézard, Lenka Zdeborová

* main: 13 pages, 5 figures; appendix: 52 pages, 4 figures 

  Access Paper or Ask Questions

Adversarial Robustness by Design through Analog Computing and Synthetic Gradients


Jan 06, 2021
Alessandro Cappelli, Ruben Ohana, Julien Launay, Laurent Meunier, Iacopo Poli, Florent Krzakala


  Access Paper or Ask Questions

Hardware Beyond Backpropagation: a Photonic Co-Processor for Direct Feedback Alignment


Dec 11, 2020
Julien Launay, Iacopo Poli, Kilian MĂĽller, Gustave Pariente, Igor Carron, Laurent Daudet, Florent Krzakala, Sylvain Gigan

* 6 pages, 2 figures, 1 table. Oral at the Beyond Backpropagation Workshop, NeurIPS 2020 

  Access Paper or Ask Questions

Epidemic mitigation by statistical inference from contact tracing data


Sep 20, 2020
Antoine Baker, Indaco Biazzo, Alfredo Braunstein, Giovanni Catania, Luca Dall'Asta, Alessandro Ingrosso, Florent Krzakala, Fabio Mazza, Marc Mézard, Anna Paola Muntoni, Maria Refinetti, Stefano Sarao Mannelli, Lenka Zdeborová

* 21 pages, 7 figures 

  Access Paper or Ask Questions

Asymptotic Errors for Teacher-Student Convex Generalized Linear Models (or : How to Prove Kabashima's Replica Formula)


Jul 01, 2020
Cedric Gerbelot, Alia Abbara, Florent Krzakala

* 15 pages main text and references, 23 pages supplementary material, 2 figures 

  Access Paper or Ask Questions

The Gaussian equivalence of generative models for learning with two-layer neural networks


Jun 25, 2020
Sebastian Goldt, Galen Reeves, Marc Mézard, Florent Krzakala, Lenka Zdeborová

* The accompanying code for this paper is available at https://github.com/sgoldt/gaussian-equiv-2layer 

  Access Paper or Ask Questions

Direct Feedback Alignment Scales to Modern Deep Learning Tasks and Architectures


Jun 23, 2020
Julien Launay, Iacopo Poli, François Boniface, Florent Krzakala

* 22 pages, 5 figures, 10 tables. For associated code, see https://github.com/lightonai/dfa-scales-to-modern-deep-learning 

  Access Paper or Ask Questions

Reservoir Computing meets Recurrent Kernels and Structured Transforms


Jun 12, 2020
Jonathan Dong, Ruben Ohana, Mushegh Rafayelyan, Florent Krzakala


  Access Paper or Ask Questions

Complex Dynamics in Simple Neural Networks: Understanding Gradient Flow in Phase Retrieval


Jun 12, 2020
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborová

* 9 pages, 5 figures + appendix 

  Access Paper or Ask Questions

Generalization error in high-dimensional perceptrons: Approaching Bayes error with convex optimization


Jun 11, 2020
Benjamin Aubin, Florent Krzakala, Yue M. Lu, Lenka Zdeborová

* 11 pages + 45 pages Supplementary Material / 5 figures 

  Access Paper or Ask Questions

Dynamical mean-field theory for stochastic gradient descent in Gaussian mixture classification


Jun 10, 2020
Francesca Mignacco, Florent Krzakala, Pierfrancesco Urbani, Lenka Zdeborová

* 8 pages + appendix, 4 figures 

  Access Paper or Ask Questions

Phase retrieval in high dimensions: Statistical and computational phase transitions


Jun 09, 2020
Antoine Maillard, Bruno Loureiro, Florent Krzakala, Lenka Zdeborová

* 11 pages (main text and references), 26 pages of supplementary material 

  Access Paper or Ask Questions

Light-in-the-loop: using a photonics co-processor for scalable training of neural networks


Jun 03, 2020
Julien Launay, Iacopo Poli, Kilian MĂĽller, Igor Carron, Laurent Daudet, Florent Krzakala, Sylvain Gigan

* 2 pages, 1 figure 

  Access Paper or Ask Questions

TRAMP: Compositional Inference with TRee Approximate Message Passing


Apr 03, 2020
Antoine Baker, Benjamin Aubin, Florent Krzakala, Lenka Zdeborová

* Source code available at https://github.com/sphinxteam/tramp. For some examples, see https://github.com/benjaminaubin/tramp_examples 

  Access Paper or Ask Questions

Double Trouble in Double Descent : Bias and Variance(s) in the Lazy Regime


Apr 03, 2020
Stéphane d'Ascoli, Maria Refinetti, Giulio Biroli, Florent Krzakala

* 29 pages, 12 figures 

  Access Paper or Ask Questions

The role of regularization in classification of high-dimensional noisy Gaussian mixture


Feb 26, 2020
Francesca Mignacco, Florent Krzakala, Yue M. Lu, Lenka Zdeborová

* 8 pages + appendix, 6 figures 

  Access Paper or Ask Questions

Generalisation error in learning with random features and the hidden manifold model


Feb 21, 2020
Federica Gerace, Bruno Loureiro, Florent Krzakala, Marc Mézard, Lenka Zdeborová


  Access Paper or Ask Questions

Asymptotic errors for convex penalized linear regression beyond Gaussian matrices


Feb 11, 2020
CĂ©dric Gerbelot, Alia Abbara, Florent Krzakala

* 31 pages, 2 figures 

  Access Paper or Ask Questions

Rademacher complexity and spin glasses: A link between the replica and statistical theories of learning


Dec 05, 2019
Alia Abbara, Benjamin Aubin, Florent Krzakala, Lenka Zdeborová

* 15 + 10 pages 

  Access Paper or Ask Questions

Exact asymptotics for phase retrieval and compressed sensing with random generative priors


Dec 04, 2019
Benjamin Aubin, Bruno Loureiro, Antoine Baker, Florent Krzakala, Lenka Zdeborová

* 12+2 pages, 6 figures 

  Access Paper or Ask Questions

Kernel computations from large-scale random features obtained by Optical Processing Units


Dec 02, 2019
Ruben Ohana, Jonas Wacker, Jonathan Dong, SĂ©bastien Marmin, Florent Krzakala, Maurizio Filippone, Laurent Daudet

* 5 pages, 3 figures, submitted to ICASSP 2020 

  Access Paper or Ask Questions

Modelling the influence of data structure on learning in neural networks


Sep 25, 2019
Sebastian Goldt, Marc Mézard, Florent Krzakala, Lenka Zdeborová


  Access Paper or Ask Questions

Who is Afraid of Big Bad Minima? Analysis of Gradient-Flow in a Spiked Matrix-Tensor Model


Jul 22, 2019
Stefano Sarao Mannelli, Giulio Biroli, Chiara Cammarota, Florent Krzakala, Lenka Zdeborová

* 9 pages, 4 figures + appendix 

  Access Paper or Ask Questions

Dynamics of stochastic gradient descent for two-layer neural networks in the teacher-student setup


Jun 18, 2019
Sebastian Goldt, Madhu S. Advani, Andrew M. Saxe, Florent Krzakala, Lenka Zdeborová

* 10 pages + references + supplemental material. arXiv admin note: substantial text overlap with arXiv:1901.09085 

  Access Paper or Ask Questions