Data is the most important part of machine learning. I believe that data should, in most cases, stay decentralized. Firstly, it avoids consolidating monopolies that thrive on early data collection. Secondly, avoiding raw data leakage is still the most important privacy feature. Thirdly, keeping data in its local context is the best way to explore better data exploitation and empower users with their data. My PhD work explored the consequences of decentralization in terms of privacy guarantees, showing that privacy attacks are still possible in this setting, but that differential privacy guarantees implemented at the scale of the node are amplified by decentralization, depending on the graph and algorithm used. More generally, I am interested in the point of view of the user in the federated and decentralized setting: what are the consequences of participating in a training? Under which settings and parameters?
Join the group
I am lucky to be part of a great team, MILES: feel free to check it out. Postdoc positions regularly open at the PSL level. At the faculty level, we also regularly hire assistant professors (maîtres de conférences), and, like me, you can join the team with a CNRS position. At the PhD level, I plan to hire one or two PhD students per year, so that the group grows slowly to a stable size where I still have time for everyone.
Current supervision
- Uddalak Mukherjee, PhD student co-supervised with Yann Chevaleyre (starting September 2026)
Previous supervision
- Lila Hammache (2026): first-year Master's internship, leading to a poster accepted at the WiML, AI4GOOD and PriLOM NeurIPS workshops
- Alireza Mirrokni (2025): ISTernship in Marco Mondelli's group, leading to an ICML paper
- Mihaela Hudişteanu (2025): ISTernship, leading to the runner-up Best Poster Award at the EurIPS PPML workshop (awarded a Sophie Germain scholarship)
- Abdellah El Mrini (2023): Master’s thesis, leading to an ICML paper (now a PhD student at EPFL)