Associate Professor, Télécom Paris
1 paper at NeurIPS 2025
We demonstrate the stability of Langevin diffusion and use it to derive the first proof of convergence for Proximal Stochastic Gradient Langevin Algorithm in a non-convex setting.