Awarded ANR JCJC funding in 2022 for the FATE project (Frugal and adaptive testing), supporting a PhD and a post-doc on bandits and sequential testing.
Co-organized the 'Complex Feedback in Online Learning' workshop at ICML 2022.
Contributing to the formalization of the polynomial Freiman-Ruzsa conjecture in Lean, led by Terence Tao.
Published numerous papers at top-tier conferences including NeurIPS, ICML, AISTATS, COLT, ALT, and CPP on topics such as best-arm identification, Markov decision processes, non-parametric bandits, and formal verification.
Formalized Doob's martingale convergence theorems in Mathlib (CPP 2023).