Greetings ! I am a PhD student at the Mila, Quebec AI Institute affiliated with Université de Montréal (UdeM), under the supervision of Ioannis Mitliagkas. I am also a part-time visiting researcher at ServiceNow Research under the supervision of João Monteiro.

Before starting my PhD, I was a visiting scholar for a year at UC Berkeley’s EECS department under the supervision of Alexandre Bayen, where I worked in collaboration with SafelyYou to detect falls in elderly care facilities. During our pilot, we were able to reduce ER visits by 80% and fall frequency by 40%.

I completed my Bachelors in computer sciences and Masters in applied mathematics at École Normale Supérieure de Paris-Saclay. I prepared for the entrance competition at Lycée Fénelon.

My research focuses on first-order optimization for deep learning. In particular, I am interested in improving our understanding of neural networks loss landscapes in a way that allows us to obtain theoretical results with practical significance. I am also interested in broader learning dynamics theory, such as understanding the relationship between the inner dimension of learned manifolds and scaling laws, multiple descent, phase transitions. Finally, I am interested in developping practical OOD detection systems to improve the reliability of predictions in real-world systems.

On top of my research, I am a co-organizer for the 4th Neural Scaling Laws Workshop, have been a teaching assistant for UdeM’s IFT3395 and IFT6390, HEC Montreal’s MATH80629A and Edulib’s SD1FR MOOC. I also co-supervised the internship of three students from Mila’s professional masters program.

My favorite paper is Stopping GAN Violence: Generative Unadversarial Networks, which I believe to be highly relevant to anyone enjoying british humor and who can afford to waste a bit of time.

Outside of work, I am fond of outdoor adventures, music festivals, and long term travels.