profile photo

Lucas Lévy

I am a first-year PhD student in the SIERRA team between INRIA Paris and ENS Ulm, where I am fortunate to work under the supervision of Francis Bach and Michael Jordan. My research interests lie at the intersection of Optimization, Machine Learning Theory and Anytime-Valid Inference.

I recently earned my Engineering Degree from Ecole polytechnique (X 2022), specializing in Applied Mathematics, alongside the MVA (Mathematics, Vision, Learning) Master's degree at ENS Paris-Saclay. Have a look at my CV for additional information.

Email / Scholar / Linkedin / Github

Publications

Beyond Optimal Rates in Stochastic Optimization: Trajectory-Adaptive Stopping Rules
Self-Concordant Perturbations for Linear Bandits
Conference on Learning Theory (COLT), 2026

Presentations

Beyond Optimal Rates in Stochastic Optimization: Trajectory-Adaptive Stopping Rules

09/2026
Master thesis defense: Trajectory-Adaptive Stopping Rules for SGD under Strong Convexity. slides

Self-Concordant Perturbations for Linear Bandits

07/2026
Invited talk: Sampling-Based Regularisation for Adversarial Linear Bandits, IMS Annual Meeting 2026, Salzburg, Austria. slides
07/2026
06/2026
12/2025
Poster, ICSDS 2025, Seville. poster