Dvir Samuel joined NVIDIA Research as a Research Scientist in 2026. His main fields of interest are machine learning and generative modeling. In particular, he studies diffusion- and flow-matching methods for image and video generation and editing, with an emphasis on personalization, controllability, and learning under long-tailed data regimes.
Dvir completed his Ph.D. at Bar-Ilan University under the supervision of Prof. Gal Chechik. His research spans long-tail and few-shot learning, as well as modern generative approaches for visual content creation and manipulation.
