Grégoire Mourre

Research interests:

• Flow-based models (diffusion models, flow matching, consistency models, flow map models, mean flow models, etc.)
• Distribution-matching models (GANs, drifting models, Wasserstein flow models, etc.)

I’m a PhD student working on generative modeling, split between the ARCHES project-team at Inria Paris and the Criteo AI Lab (CAIL, FDL team) in Paris. I am supervised by Claire Monteleoni, Emmanuel de Bézenac, Jean-Yves Franceschi and Thibaut Issenhuth.

I’m especially interested in one-step and few-step generative models for fast inference, and I care about both the underlying theory and the implementation tricks that make these methods actually work.

Before starting my PhD, I spent time as a deep learning research intern at Dassault Systèmes working on generative world models. Outside of research, you’ll usually find me bouldering, running, at the cinema, or losing at board games.