Publications

2025

  1. Preprint
    Asymptotics of SGD in Sequence-Single Index Models and Single-Layer Attention Networks
    Luca Arnaboldi, Bruno Loureiro, Ludovic Stephan, Florent Krzakala, and Lenka Zdeborova
    arXiv preprint, Jun 2025
    arXiv

2024

  1. Preprint
    Community detection with the Bethe-Hessian
    Ludovic Stephan and Yizhe Zhu
    arXiv preprint, Nov 2024
    arXiv
  2. Journal
    How Two-Layer Neural Networks Learn, One (Giant) Step at a Time
    Yatin Dandi, Florent Krzakala, Bruno Loureiro, Luca Pesce, and Ludovic Stephan
    Journal of Machine Learning Research, Nov 2024
    HTML PDF
  3. Conference
    Online Learning and Information Exponents: The Importance of Batch size & Time/Complexity Tradeoffs
    Luca Arnaboldi, Yatin Dandi, Florent Krzakala, Bruno Loureiro, Luca Pesce, and Ludovic Stephan
    In Proceedings of the 41st International Conference on Machine Learning, Jul 2024
    HTML PDF
  4. Conference
    A non-backtracking method for long matrix and tensor completion
    Ludovic Stephan and Yizhe Zhu
    In Proceedings of Thirty Seventh Conference on Learning Theory, Jun 2024
    HTML PDF Slides
  5. Preprint
    Repetita iuvant: Data repetition allows sgd to learn high-dimensional multi-index functions
    Luca Arnaboldi, Yatin Dandi, Florent Krzakala, Luca Pesce, and Ludovic Stephan
    arXiv preprint, May 2024
    arXiv
  6. Journal
    Gaussian universality of perceptrons with random labels
    Federica Gerace, Florent Krzakala, Bruno Loureiro, Ludovic Stephan, and Lenka Zdeborová
    Physical Review E, Mar 2024
    arXiv HTML Poster
  7. Journal
    Sparse random hypergraphs: non-backtracking spectra and community detection
    Ludovic Stephan and Yizhe Zhu
    Information and Inference: A Journal of the IMA, Feb 2024
    Also in FOCS’22
    DOI arXiv HTML Slides

2023

  1. Conference
    Universality laws for Gaussian mixtures in generalized linear models
    Yatin Dandi, Ludovic Stephan, Florent Krzakala, Bruno Loureiro, and Lenka Zdeborová
    In Advances in Neural Information Processing Systems, Dec 2023
    HTML
  2. Conference
    Are Gaussian Data All You Need? The Extents and Limits of Universality in High-Dimensional Generalized Linear Estimation
    Luca Pesce, Florent Krzakala, Bruno Loureiro, and Ludovic Stephan
    In Proceedings of the 40th International Conference on Machine Learning, Jul 2023
    HTML PDF Poster
  3. Conference
    From high-dimensional & mean-field dynamics to dimensionless odes: A unifying approach to sgd in two-layers networks
    Luca Arnaboldi, Ludovic Stephan, Florent Krzakala, and Bruno Loureiro
    In The Thirty Sixth Annual Conference on Learning Theory, Jul 2023
    HTML PDF
  4. Preprint
    Escaping mediocrity: how two-layer networks learn hard single-index models with SGD
    Luca Arnaboldi, Florent Krzakala, Bruno Loureiro, and Ludovic Stephan
    arXiv preprint, May 2023
    arXiv

2022

  1. Journal
    Non-backtracking spectra of weighted inhomogeneous random graphs
    Ludovic Stephan and Laurent Massoulié
    Mathematical Statistics and Learning, Dec 2022
    DOI HTML PDF
  2. Conference
    Phase diagram of Stochastic Gradient Descent in high-dimensional two-layer neural networks
    Rodrigo Veiga, Ludovic Stephan, Bruno Loureiro, Florent Krzakala, and Lenka Zdeborová
    In Advances in Neural Information Processing Systems, Dec 2022
    HTML PDF Poster

2021

  1. Preprint
    A simpler spectral approach for clustering in directed networks
    Simon Coste and Ludovic Stephan
    arXiv preprint, Feb 2021
    arXiv

2019

  1. Conference
    Planting trees in graphs, and finding them back
    Laurent Massoulié, Ludovic Stephan, and Don Towsley
    In Conference on Learning Theory, Jun 2019
    HTML PDF
  2. Conference
    Robustness of Spectral Methods for Community Detection
    Ludovic Stephan and Laurent Massoulié
    In Conference on Learning Theory, Jun 2019
    HTML PDF