• I am Machine Learning Scientist

    I am currently leading the ML research team at Init ML working on Computer Vision tasks such as matting, inpainting, super-resolution, depth and surface-normals estimation for our ClipDrop products.

  • I am Ph.D in Statistical Physics and Machine Learning



    I recently obtained my Ph.D from IPhT (CEA Saclay) and ENS Paris under the supervision of Lenka Zdeborová and Florent Krzakala. I worked on the statistical physics of disordered systems, information theory, statistical inference and message-passing algorithms and their applications to theoretical machine learning (neural networks, phase retrieval, compressed sensing, matrix factorization,...).

    The manuscript and the slides are available at this address and this address.


    View my research projects

  • I also interned at Facebook Artificial Intelligence Research

    Last year, I started a new research direction at Facebook AI Research on causal and invariant features learning with Léon Bottou and David Lopez-Paz

    I investigated how to degsign alternatives to gradient-based algorithms to perform Empirical Risk Minimization that automatically disregard spurious features and rely only on invariant and robust ones to generalize out-of-distribution.

    We especially proposed six linear low-dimensional problems -- unit tests -- to evaluate different types of out-of-distribution generalization described in this paper

Who am I?

I have recently obtained my Ph.D from Institut de Physique Théorique (IPhT - CEA Saclay, Université Paris-Saclay). Prior to this, I obtained a Master of Science in Theoretical physics at Ecole Normale Supérieure de Paris and a Master of Science and Engineering from Ecole polytechnique. I obtained as well a Bachelor of Science from Ecole polytechnique and was student in Classe Préparatoire aux Grandes Ecoles at Lycee Henri IV in Paris.

I am interested in statistical physics of disordered systems and I worked during my Ph.D with Lenka Zdeborová and Florent Krzakala on the application of these methods to provide a theoretical understanding of deep neural networks and more classical machine learning models and algorithms.

In my spare time, I used to play rugby 🏈 and I still practise skydiving, windsurfing, kitesurfing, surfing 🏄, hiking, biking 🚵‍♂️ and a lot of sport in general!

Research projects

Linear unit-tests for invariance discovery

NeurIPS 2020 workshop - Causal Discovery & Causality-Inspired Machine Learning February 22, 2021

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

NeurIPS 2020 June 12, 2020

Compositional Inference with TRee Approximate Message Passing

Submitted to Journal of Machine Learning Research

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

MSML 2020 December 5, 2019

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

MSML 2020 September 14, 2019

The spiked matrix model with generative priors

NeurIPS 2019 June 1, 2019

Storage capacity of binary symmetric perceptrons

Journal of Physics A: Mathematical and Theoretical Janvier 2, 2019

The committee machine: Computational to statistical gaps in learning a two-layers neural network

NeurIPS 2018 June 14, 2018

Personal projects

A simple webservice and mobile app for Natural Language Understanding

NLU, frontend web and app

Background removal service

Matting with a trimap

Summary of my skills

Statistical physics

90%

Python, Matlab

90%

Machine learning and deep learning theory

90%

PyTorch, Scikit-Learn, numpy, pandas, Tensorflow

90%

Bash, git, Docker, Kubernetes, GCP, AWS

80%

HTML5, CSS3, JS, PHP, SQL

60%

React, ReactNative

70%

Java, C++, Rust, Julia

20%

Theoretical physics

Machine learning

Frontend

Backend

DevOps / MLOps

Education

Statistical physics of disordered systems, Constraints statisfaction problems, Machine Learning, Deep Learning
Statistical physics, Numerical physics, Bayesian inference, Statistical field theory, Condensed matter, Non-linear physics, Stochastic processes, Non-equlibirum physics
Advanced quantum mechanics, Statistical physics, Computer science (advanced programming and algorithms), Probability, Statistics, Massive data processing, Control dynamic models,Fluid mechanics, Continuum mechanics, Particle physics, Soft mmatter, Complex systems, Quantum optics.
Mathematics, Physics, Chemistry, Litterature, Foreign languages

Work Experience

Init ML - ClipDrop March 2021 -

Lead machine learning scientist @ ClipDrop, computer vision deep models for matting, inpainting, super-resolution, depth and surface-normals estimation, text-to-image generative models, ...

Facebook Artificial Intelligence Research March - December 2020

As suggested by the formulation "most cows appear in grass and most camels appear in sand", Empirical Risk Minimization relies essentially on the dominant background color pixels to produce a prediction. To address this issue, the idea is to learn invariant features across multiple training distributions and to use those correlations as a proxy for out-of distribution generalization. To clarify the situation, as a core project we started to design simple, memory cheap, linear 'unit-test' that already capture OoD generalization failures. We designed and proposed three linear problems (with three scrambled versions) that contain invariant causal correlations that we would like to learn, as well as a spurious correlation that we would like to discard.

Ph.D, Université Paris-Saclay October 2017 - December 2020

I established the phase diagrams of various machine learning models, using the replica method and message passing algorithms from statistical physics of disordered systems, with an emphasis on the potential differences between statistical and algorithmic thresholds, for simple synthetic and theoretically tractable tasks.

Research Internship, Massachusetts Institute of Technology (MIT) April 2016 - August 2016

Yves Couder, Emmanuel Fort and coworkers recently discovered that a millimetric droplet sustained on the surface of a vibrating fluid bath may self-propel through a resonant interaction with its own wave field. This internship within the department of Mathematics, aimed to study such a droplet, called a walker, crossing a submerged obstacle. This difference of depth, due to the obstacle, is analog to a changement of media and optical index in optics and thus we showed a Snell-Descartes law for walking droplets.

Research project, Université Paris-Sud, Laboratoire de Physique des Solides April 2016 - August 2016

The advent of graphene and topological insulators in 2005 and 2006 was a revolution in condensed matter physics. The main specificity of these two-dimensional materials have probed some theories established in very remote contexts in condensed matter, such as quantum mechanics and relativistic field theory, rather encountered in high-energy physics. It is about understanding the appearance of topological band structures in two-dimensional materials, semiconductors or semi-metal, covered by models called " strong bonds " and the study of electrical and thermal transport properties these particular materials by means of statistical physics .

Research internship, Thales Research and Technology April 2016 - August 2016

I was responsible of developing a measuring bench of the influence of external disturbances including sensitivity to mechanical vibration of new optoelectronic oscilateurs (OEO), which allow the generation of signals with very low noise for radar applications.

Get in Touch