Pré-publication, Document de travail, AO, Mécanique statistique

Designing losses for data-free training of normalizing flows on Boltzmann distributions

Loris Felardos, Jérôme Hénin, Guillaume Charpiat. Designing losses for data-free training of normalizing flows on Boltzmann distributions. 2023. ⟨hal-03936982⟩

Publié le