Profile image

Edwige Cyffers

Permanent CNRS Researcher at Paris Dauphine-PSL

Master's internships and PhD positions available! (see the bottom of this page)

Since 2026, I have been a permanent CNRS researcher (chargée de recherche, a tenured position) at LAMSADE, Paris Dauphine-PSL, where I am starting my own group as PI. If you would like to do a PhD under my supervision, see below! This year, I will teach the Cours Peccot and serve as NeurIPS Satellite Chair for Paris.

I was a Research Fellow during the Federated Learning semester at the Simons Institute in Berkeley. Previously, I was a postdoc at ISTA working with Christoph Lampert and Marco Mondelli. My PhD was supervised by Aurélien Bellet in the Magnet team, on the topic of Differential Privacy for Decentralized Learning, and I was awarded the L’Oréal-UNESCO Young Talents award, the SSFAM PhD award, and the runner-up AFIA PhD award for my doctoral work. Before that, I studied Computer Science at ENS Lyon as a civil servant. I graduated as a civil engineer through a dual-degree programme, and I hold a Master’s degree in applied mathematics for machine learning (MVA) and a Master’s degree in Philosophy. I have worked on various science popularization projects, and I hope to keep exploring the philosophical and outreach aspects of machine learning. I am particularly drawn to privacy, fairness, accountability, and regulatory issues in ML from a mathematical point of view.

As a woman in STEM, and more particularly in CS/maths research, I am a firm believer in diversity, and I strive to advocate for it at a personal level by volunteering in projects that promote girls and women in STEM. I take part in many local initiatives and would be happy to join yours! However, if you are looking for a role model, it is now well established that the most effective role models are those closest to the target audience (e.g., undergraduates for high school students, Master's or PhD students for undergraduates). Please also keep in mind that my time and money are limited.

I organized EGMO 2026 in France. The European Girls’ Mathematical Olympiad is a very high-level international mathematics competition that aims to encourage girls to discover and explore their mathematical talent.

How to pronounce my name?

I don't mind if your pronunciation is unusual, and I am very bad at pronouncing foreign names myself. You can pronounce my first name the way Harry pronounces his owl's name in Harry Potter. For my last name, "Il faut Cyffers" and "Il faut s'y faire" sound the same in French, and "Cyffers" sounds nearly the same as "see fair".

Join me

I am hiring Master’s interns, with the possibility of continuing with a PhD afterwards; the funding is already secured. Please also read my research page.

Here are some possible topics:

  • Performative learning datasets: Description
  • Exploring the properties of DP algorithms (such as DP-SGD) with tools from ML theory
  • Open topic: if you already have a topic in mind, feel free to propose it. This is how my own PhD started, and it is entirely fine if you want to work with me but do not like the offers above.

Last year, I completely failed to answer, or even read with proper care, all the emails I received. So this year I am trying something new: I will recruit through the ELLIS PhD program, so please apply there! If you contact me by email (for instance, if you are at ENS or Polytechnique and want an internship while your 4A is still ahead of you), please follow these guidelines.

The email subject line should be: [Internship application] + the topic you are interested in. The email should include what you hope to achieve by doing a PhD, what you can do in maths and in code, which topics interest you among Privacy, Performative Learning, Fairness, Unlearning, Federated Learning, and Decentralized Learning, and optionally the name of a professor who knows you. The whole email should be at most 500 words. I speak French and English.

Research Interests

  • Privacy in ML
  • Federated and Decentralized ML
  • Trustworthy ML

Education

  • PhD, 2021-2024
    Inria/Université de Lille
  • MSc MVA, 2019-2020
    PSL/ENS Paris-Saclay
  • MA LOPHISC, 2019-2021
    Sorbonne University
  • MSc Civil Engineer, 2020-2021
    Mines de Paris
  • Normalienne, 2017-2021
    ENS Lyon