Arnak Dalalyan, a statistics specialist at CREST, receives an ERC Advanced Grant

Arnak Dalalyan's research in statistics focuses on its application to machine learning, with particular emphasis on robust methods, high-dimensional data and sampling techniques. He is also a professor at ENSAE and a fellow at the interdisciplinary center Hi! Paris.
The ERC Advanced will fund his project SAGMOS (Statistical Analysis of Generative Models: Sampling Guarantees and Robustness).
Generative artificial intelligence models - algorithms capable of creating realistic texts, images, music or molecular structures - are at the heart of many technological innovations today. The SAGMOS project will provide new mathematical guarantees on the reliability, originality and efficiency of these models, in particular the diffusion models that have become the new standard in this field.
Read the article on the CREST website.
*CREST: a joint research unit CNRS, École Polytechnique, GENES, ENSAE Paris, Institut Polytechnique de Paris, 91120 Palaiseau, France