Valentin De Bortoli is a research scientist at CNRS and the Center for Data Science in Ecole Normale Supérieure in Paris. His research lies at the intersection between applied probability, statistics and machine learning with a recent focus on the interplay between stochastic processes, optimal transport, control and diffusion models. His long-term objective is to provide clear theoretical guarantees for generative models and to foster their applications to other areas of machine learning. Prior to joining CNRS, he was a postdoctoral researcher at Oxford University working with Arnaud Doucet working on diffusion models. Valentin received his Ph.D from Ecole Normale Supérieure in 2020, under the supervision of Agnes Desolneux, Bruno Galerne and Arthur Leclaire. During his Ph.D he developed new algorithms for texture synthesis based on information geometric approaches similar to energy-based models. He also worked on the theory of Markov chains and their continuous-time counterparts.

---