Scientific activities

Seminars

The CSD seminar programme and archive: research talks in data science and artificial intelligence.

Salle Dussane at ENS, Paris

For 2026–2027, seminars usually take place on Thursdays, 11:00–12:00. Tuesday colloquia usually take place at 11:00–12:00. Check each event for exceptions. All times are Paris local time.

Upcoming seminars

11:00 · Room W (default room)
UpcomingTo be confirmedSeminar

Nicolas Schreuder (CNRS / École normale supérieure, Paris)

-

11:00 · Room W (default room)
UpcomingSeminar

Jerome Garnier Brun (Bocconi University)

-

11:00 · Room W (default room)
UpcomingColloquium

Marc Lelarge (Inria / École normale supérieure, Paris)

-

11:00 · Room W (default room)
UpcomingSeminar

Xiaoming Yuan (University of Hong Kong)

-

11:00 · Room W (default room)
UpcomingColloquium

Pierre Gentine (Columbia University)

-

11:00 · Room W (default room)
UpcomingTo be confirmedSeminar

Antonin de Courcel (École normale supérieure, Paris)

Effective models in high dimensional unstable regimes: optimization dynamics at the edge of stability

11:00 · Room W (default room)
UpcomingColloquium

Alexander Mathis (EPFL)

-

14:00 · Room W (default room)
UpcomingTo be confirmedSeminar

Steven L. Brunton (University of Washington)

-

CSD · ENS–PSL

Seminar archive

Organiser sign-in

127 talks

11:00
Seminar

Samuel Vaiter (CNRS / Université Côte d’Azur)

(Trying to) Understand Steering in Language Models

11:00
Working group

Vladimir Petrovic; Damien Rouchousse

Fast coresets for the Sinkhorn divergence / Discrete Diffusion Design: An Asymptotic Error Analysis Approach

Talks in this session

Vladimir Petrovic · Fast coresets for the Sinkhorn divergence

Damien Rouchousse · Discrete Diffusion Design: An Asymptotic Error Analysis Approach

11:00
Working group

Guy Gilboa (Technion – Israel Institute of Technology)

How to Encode World Knowledge?

12:00
Seminar

Florent Krzakala (EPFL)

Deep Learning as Neural Low-Degree Filtering: A Spectral Theory of Hierarchical Feature Learning

12:00
Seminar

Emmanuel Sérié; Gabriel Peyré

LLM for coding tutorial

Archive note

The document marks Gabriel Peyré’s participation as provisional.

12:00
Seminar

Constant Bourdrez; Hanna Benarroch (École normale supérieure, Paris)

Learning To Sample From Diffusion Models Via Inverse Reinforcement Learning / Certified Per-instance Unlearning using Individual Sensitivity Bounds

Talks in this session

Constant Bourdrez · Learning To Sample From Diffusion Models Via Inverse Reinforcement Learning

Hanna Benarroch · Certified Per-instance Unlearning using Individual Sensitivity Bounds

12:00
Seminar

Théo Lacombe (Université Gustave Eiffel)

Learning Monge Maps with Constrained Drifting Models

Colloquium

Rebecca Willet (University of Chicago)

How do simple rotations affect the implicit bias of Adam?

Seminar

Fabian Schaipp (Inria / École normale supérieure, Paris)

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Seminar

Jiayi Li (Max Planck Institute of Molecular Cell Biology and Genetics)

A Basin-Selection Perspective on Grokking via Singular Learning Theory

Seminar

Valérie Castin; Gauthier Thurin (École normale supérieure, Paris)

Balanced Low-Rank Adaptation: Removing Parameter Invariance to Accelerate Convergence / Convergence Rates for Distribution Matching with Sliced Optimal Transport

Talks in this session

Valérie Castin · Balanced Low-Rank Adaptation: Removing Parameter Invariance to Accelerate Convergence

Gauthier Thurin · Convergence Rates for Distribution Matching with Sliced Optimal Transport

CancelledSeminar

Eugène Ndiaye (Apple)

Beyond Uncertainty Sets: Leveraging Optimal Transport to Extend Conformal Predictive Distribution to Multivariate Settings

Seminar

Batiste Le bars (Inria Lille)

Conformal Prediction for Uncertainty Quantification in Machine Learning: Recent Advances

Colloquium

Michael Chertkov (University of Arizona)

Samples That Cooperate, Samples That Remember: Two Exactly Solvable Bridge Diffusions

Seminar

Tony Silveti-Falls (CentraleSupélec / Université Paris-Saclay)

Training Neural Networks at Any Scale

Seminar

François Malgouyres (Institut de Mathématiques de Toulouse)

Geometry-induced regularization and identifiability of deep ReLU networks

Seminar

Eugène Ndiaye (Apple)

-

Colloquium

Julien Mairal (Inria Grenoble)

Physical Models and Machine Learning for Scientific Imaging

Working group

Leonardo Defilippis; Luigi Fogliani (École normale supérieure, Paris)

Scaling Laws and Spectra of Shallow Neural Networks in the Feature Learning Regime / Annealing in variational inference mitigates mode collapse: A theoretical study on Gaussian mixtures

Talks in this session

Leonardo Defilippis · Scaling Laws and Spectra of Shallow Neural Networks in the Feature Learning Regime

Luigi Fogliani · Annealing in variational inference mitigates mode collapse: A theoretical study on Gaussian mixtures

Seminar

Thibaut Germain (École polytechnique)

A Spectral-Grassmann Wasserstein metric for operator representations of dynamical systems

Colloquium

Alessandro Laio (SISSA)

Identifying informative distance measures in high-dimensional feature spaces

Working group

Nina Vesseron; Johannes Hertrich (Nina Vesseron: CREST / ENSAE Paris; Johannes Hertrich: Université Paris Dauphine-PSL / Inria / ENS)

On a Neural Implementation of Brenier's Polar Factorization / On the Relation between Rectified Flows and Optimal Transport

Talks in this session

Nina Vesseron · On a Neural Implementation of Brenier's Polar Factorization

Johannes Hertrich · On the Relation between Rectified Flows and Optimal Transport

Seminar

Ashok Vardhan (Télécom Paris)

From Markov to Laplace: A Markovian Tale of Large Language Models

Abstract

Large Language Models based on transformers and state-space models (Mamba) are at the forefront of recent breakthroughs in a variety of disciplines, including natural languages. Despite their remarkable empirical success, our theoretical understanding of how these models learn and represent sequential structure, and how in-context learning (ICL) capabilities emerge on such tasks, remains limited. This raises a fundamental question: How do modern sequence models learn from sequential data? To address this question, in this talk I will present our new framework for a principled theoretical and empirical analysis of LLMs via Markov chains. The key idea underpinning our approach is the modeling of sequential input data as a Markov process, inspired by the Markovianity of natural languages. We utilize this framework to systematically study the interplay between the Markov order and model depth, revealing sharp contrasts between transformers and state-space models. In particular, our analysis reveals the curious phenomena that (i) even a single-layer Mamba efficiently learns the in-context Laplacian smoothing estimator, whereas transformers require two layers, and (ii) while a single layer transformer cannot efficiently represent a conditional $k$-gram model, a two layer transformer can represent it for any Markov order $k \geq 1$. Together, our results provide the first formal connection between Mamba and optimal statistical estimators, and yield the tightest known characterization of the interplay between transformer depth and Markov order for ICL. Deepening our understanding of LLMs and their ICL capabilities, we believe our framework provides a new avenue for a principled study of LLMs with plenty of interesting open questions abound, which I will discuss in the end.

Working group

Eddie Aamari (CNRS / École normale supérieure, Paris)

-

Seminar

Marine Le Morvan (Inria)

A Tabular Foundation Model for In-Context Learning on Large Data

Colloquium

Nathan Srebro (Toyota Technological Institute at Chicago)

Learning to Answer from Correct Demonstrations

Seminar

Pierre Marion (EPFL)

Phase Diagram of Dropout for Two-Layer Neural Networks in the Mean-Field Regime

Working group

Etienne Lempereur; Raphaël Urfin

Moment Guided Diffusion for Maximum Entropy Generation / Why Diffusion Models Don't Memorize: The Role of Implicit Dynamical Regularization in Training

Talks in this session

Etienne Lempereur · Moment Guided Diffusion for Maximum Entropy Generation

Raphaël Urfin · Why Diffusion Models Don't Memorize: The Role of Implicit Dynamical Regularization in Training

Seminar

Jérôme Bolte (Toulouse School of Economics / École normale supérieure, Paris)

Optimization Perspectives on the Differentiation of Programs

Seminar

Stephen Zhang (University of Melbourne)

Flow matching for (generalized) Schrödinger bridges and dynamical optimal transport

Colloquium

Étienne Ollion (CNRS / École polytechnique)

-

Seminar

Hugo Cui (Harvard University)

-

Working group

Samuel Boïté; Léa Bohbot

-

Working group

Arie Wortsman; Benjamin Dupuis

-

Colloquium

Eva Dyer (University of Pennsylvania)

Large-scale pretraining on neural data allows for transfer across individuals, tasks and species

Working group

Arthur Pellegrino; Nikos Tsilivis

-

Seminar

Emile Pierret (Université Paris Cité / École normale supérieure, Paris)

-

Working group

Davide Carbone; Nathanael Cuvelle-Magar

-

Seminar

Julie Delon (École normale supérieure, Paris)

Computing Barycentres of Measures for Generic Transport Costs

Seminar

Luiz Chamon (École polytechnique)

Learning under requirements

Working group

Kruno; Luca del Bono

-

Colloquium

David Louapre (Ubisoft)

-

Colloquium

Eric Vanden-Eijnden (New York University)

-

Seminar

Grant Rotskoff (Stanford University)

-

Seminar

Loucas-Pillaud Vivien (École des Ponts ParisTech)

-

Working group

Beatrice Achilli (Bocconi University)

-

Working group

Olivier Cappé; Gabriel Peyré (CNRS / École normale supérieure, Paris)

Modern Transformers and GRPO

Seminar

Denny Wu (New York University / Flatiron Institute)

-

Seminar

Demba Ba (Harvard University)

Ten years of playing in deep haystacks: a taxonomy of interpretability research

Seminar

Alex Cayco Gajic (École normale supérieure, Paris)

-

Seminar

Yann Fleureau (Project Numina)

Projet Numina: IA appliquée au raisonnement et à la formalisation mathématiques

Seminar

Antonio Ocello (École polytechnique)

-

Seminar

Clément Bonet (CREST / ENSAE Paris)

-

Colloquium

Fanny Yang (ETH Zürich)

-

Seminar

Maxence Noble (École polytechnique)

-

Seminar

Alessandro de Palma (Inria Paris)

-

Colloquium

Michele Ceriotti (EPFL)

-

Working group

Antonin Brossollet; Etienne Lempereur

-

Working group

Gauthier Thurin; Leonardo Defilippis (CNRS / École normale supérieure, Paris)

-

Working group

Olivier Cappé (CNRS / École normale supérieure, Paris)

-

Working group

Emanuele Francazi; Alexis Aymé

-

Colloquium

Jean-Remi King (CNRS / École normale supérieure, Paris / Meta AI)

-

Working group

Carl Allen (École normale supérieure, Paris)

-

Seminar

Elisabetta Cornacchia (Inria / École normale supérieure, Paris)

-

Colloquium

Luca Biferale (University of Rome Tor Vergata)

-

Seminar

Nicolas Schreuder (LIGM)

Fairness in machine learning : a study of the Demographic Parity constraint

Archive note

The document’s “Nicolas Scwhreuder” spelling was reconciled with the matching date and full name in the old public archive.

Working group

Louis Grenioux; Christoph Schönle

-

Working group

Alexandre Vérine; Jérémie Kalfon

-

Seminar

Gabriele Sicuro (Bologna)

The random assignment problem

Working group

Tony Bonnaire (École normale supérieure, Paris / Université PSL)

-

Seminar

Vassilis Papadopoulos (EPFL)

Arrows of Time for Large Language Models

Seminar

Matthieu Wyart (EPFL)

Learning hierarchical representations with deep nets and large language models

Seminar

Carl Allen (ENS)

Variational Classification

Working group

Maarten de Hoop (Rice University)

Lecture on neural operators

Archive note

The planning document records two lectures, 20 June and 2 July. The speaker’s “Martin” spelling is reconciled with the document’s Maarten de Hoop entry and the CSD visiting-professor list.

Colloquium

Andrea Montanari (Stanford University)

-

Working group

Maarten de Hoop (RICE)

Lecture on neural operators

Archive note

The planning document records two lectures, 20 June and 2 July. The speaker’s “Martin” spelling is reconciled with the document’s Maarten de Hoop entry and the CSD visiting-professor list.

Seminar

Raphael Berthier (Sorbonne)

The two timescale regime for the analysis of neural networks

Seminar

Ehsan Elhamifar (Northeastern)

Understanding Complex Procedural Videos

Working group

Nina Vesseron; Francisco Andrade

On a Neural Implementation of Brenier's Polar Factorization / Sparsistency for Inverse Optimal Transport

Talks in this session

Nina Vesseron · On a Neural Implementation of Brenier's Polar Factorization

Francisco Andrade · Sparsistency for Inverse Optimal Transport

Seminar

Kimia Nadjahi (ENS)

Slicing Mutual Information Generalization Bounds for Neural Networks

Colloquium

M. Jordan

-

Archive note

The document records the speaker as “M Jordan”; the name has not been expanded.

Working group

Kimia Nadjahi; Samuel Hurault

-

Seminar

Malick

-

Archive note

Only the speaker’s surname or first name “Malick” is recorded in the document.

Colloquium

Lenaic Chizat (EPFL)

-

Seminar

Ulugbek Kamilov (Washington University in St. Louis)

Computational Imaging: Restoration Deep Networks as Image Priors

Colloquium

Antoine Georges (Collège de France / Flatiron Institute)

-

12:00
Working group

Florent Krzakala (EPFL)

How do neural nets learn?

Seminar

Aymeric Dieuleveut (École polytechnique)

-

Colloquium

Noah Smith (University of Washington)

-

Seminar

Cyril Letrouit (Orsay)

Two mathematical perspectives on Transformers

Working group

Maarten de Hoop (Rice)

Advances in the design and analysis of neural operators

Seminar

Francis Bach (INRIA)

An alternative view of denoising diffusion models

Seminar

Lenka Zdeborová (EPFL)

Backtracking Dynamical Cavity method

Working group

Lenka Zdeborová (EPFL); Backtracking Dynamical Cavity method

-

Working group

Gabriel

Tutorial on Autodiff

Archive note

Only the speaker’s first name is recorded in the document.

Seminar

Mathieu Desbrun (Inria / École polytechnique)

Blue Noise Sampling

Seminar

Pierre-Alexandre Mattei (Inria Sophia Antipolis)

Are ensembles getting better All the time ?

Working group

Leonardo Defilippis (CNRS / École normale supérieure, Paris)

High-dimensional estimation under heavy-tails, with applications to robust regression

Archive note

The document’s “Leonardo Defillipis” spelling was reconciled with its other entries and the public archive.

Seminar

Antoine Maillard (ETH Zürich)

Fitting ellipsoids to random points

Working group

Etienne Lempereur; Edwige Cyffers

Learning Physics with Wavelet Renormalisation Group / Differential Privacy in Decentralized Learning

Talks in this session

Etienne Lempereur · Learning Physics with Wavelet Renormalisation Group

Edwige Cyffers · Differential Privacy in Decentralized Learning

Working group

Nathanael Cuvelle-Magar; Raja Giryes

-

12:30
Seminar

Andrea Montanari (Stanford University)

Sampling via diffusion processes : rigorous guarantees, hardness, disorder chaos

Working group

Michael Sander (CNRS / École normale supérieure, Paris)

-

Colloquium

Julia Kempe (New York University)

-

Working group

Valérie Castin

Theory of ML journal club

Seminar

Soledad Villar (Johns Hopkins University)

Machine learning and invariant theory

Seminar

Antonio Silveti-Falls (CentraleSupélec / Université Paris-Saclay)

Differentiating Nonsmooth Solutions to Parametric Monotone Inclusion Problems

Seminar

Francois Lanusse (CNRS / CEA Paris-Saclay)

Deep Generative Models for Hybrid Physical/Data-Driven Bayesian Inference

Seminar

Pierre Ablin (Apple)

Training neural networks with orthogonal weights

Seminar

Arthur Mensch

Flamingo : a Visual Language Model for Few-Shot Learning

Seminar

Stéphanie Allassonnière (Université Paris Cité / École polytechnique)

Data-Augmentation in High Dimensional Low Sample Size Setting Using a Geometry-Based Variational Autoencoder

Seminar

James Thornton

Diffusion / Field based Reconstruction and Generation for 3D Shapes

Seminar

Gaël Varoquaux (Inria)

Embeddings to learn on messy relational data

Seminar

Hervé Jégou

Learning image representations with coarse, instance-level and image-level supervision

Seminar

François Charton (Meta AI / FAIR Paris)

Deep learning for symbolic maths

Archive note

Date retained exactly as published. This entry appears after December 2021 in the source list; its year may need archival verification.

Seminar

Mathieu Wyart (EPFL)

What makes data learnable by deep learning

Archive note

Date retained exactly as published. This entry appears after December 2021 in the source list; its year may need archival verification.

Seminar

Rémi Monasson (CNRS / École normale supérieure, Paris)

Restricted Boltzmann Machines revisited : from sampling to design

Archive note

Date retained exactly as published. This entry appears after December 2021 in the source list; its year may need archival verification.

Seminar

Bruno Loureiro (CNRS / École normale supérieure, Paris)

Phase diagram of Stochastic Gradient Descent in high-dimensional two-layer neural networks

Seminar

Raja Giryes (Tel Aviv University)

Sampling based analysis of neural network generalization and extrapolation

Seminar

Francis Bach (Inria / École normale supérieure, Paris)

The quest for adaptivity

Seminar

Rachel Bawden (Inria)

Low-resource MT : few-shot learning and historical language normalisation

Seminar

Thomas Moreau (Inria Saclay)

Convolutional Sparse Coding for Electromagnetic Brain Signals

Seminar

Marylou Gabrié (École polytechnique)

Enhancing Sampling with Learning

Seminar

Charles Martin (Calculation Consulting)

A Semi-Empirical Model for the Generalization Capacity of Deep Neural Networks

Seminar

Marc Lelarge (Inria / École normale supérieure, Paris)

Exploiting Graph Invariants in Deep Learning

Seminar

Sebastian Goldt (SISSA)

The interplay of data structure and learning dynamics in simple neural networks

Seminar

Dieuwke Hupkes - On locality (Facebook AI Research)

globality, consistency and compositionality in neural machine translation

Historical entries retain the information published in the old archive. Missing titles, times, and rooms have not been reconstructed.

Download the seminar database (JSON)

CSD mailing list

Receive announcements of seminars, courses, and events at the centre.

Join the mailing list