|
Research
My PhD is about modelling, learning and sampling high dimensional maximum entropy distributions. Along with my supervisor and co-authors, we have developped frameworks using both the renormalisation group and stochastic interpolants, with applications to out of equilibrium physical fields (including active matter and turbulence), finance and real world datasets.
|
Publications
Conditionally Strongly Log-Concave Generative Models
Florentin Guth*, Etienne Lempereur*, Joan Bruna, Stéphane Mallat
Proceedings of the 40th International Conference on Machine Learning, PMLR 202:12224-12251,
paper
Hierarchic flows to estimate and sample high-dimensional probabilities
Etienne Lempereur, Joan Bruna, Stéphane Mallat
Applied and Computational Harmonic Analysis, Volume 83, 2026
paper
Effective energy, interactions and out of equilibrium nature of scalar active matter
Antonin Brossollet*, Etienne Lempereur*, Stéphane Mallat & Giulio Biroli
Communications Physics volume 8,499, 2025
paper
MGD: Moment Guided Diffusion for Maximum Entropy Generation
Etienne Lempereur, Nathanaël Cuvelle--Magar, Florentin Coeurdoux, Stéphane Mallat, Eric Vanden-Eijnden
Arxiv Preprint, 2026
paper
Training-Free Generative Modeling via Kernelized Stochastic Interpolants
Florentin Coeurdoux, Etienne Lempereur, Nathanaël Cuvelle-Magar, Thomas Eboli, Stéphane Mallat, Anastasia Borovykh, Eric Vanden-Eijnden
Arxiv Preprint, 2026
paper
|
|
