Near-Lipschitz stability of the Kim--Milman flow map
Chewi, S., Eichinger, K., and Pooladian, A-A. (alphabetical)
I am currently a Foundations of Data Science (FDS) Postdoctoral associate at Yale University. Prior to this, I had the (immense!) pleasure of completing my PhD at New York University under the supervision of Jonathan Niles-Weed. My thesis can be found here. Before my PhD, I completed my BA and MSc in Applied Mathematics at McGill University, where I worked with Tim Hoheisel and Adam Oberman.
I study statistical and mathematical aspects of data science. My recent work focuses on analyzing and developing methods for large-scale probabilistic inference and generative modeling, often through the lens of optimal transport theory.
Select Publications
All PublicationsBlind denoising diffusion models and the blessings of dimensionality
Kadkhodaie, Z.*, Pooladian, A-A.*, Chewi, S., and Simoncelli, E. (Joint first author)
Spotlight at FoGEN ICML workshop; 2026
Wasserstein Flow Matching: Generative modeling over families of distributions
Haviv, D.*, Pooladian, A-A.*, Pe'er, D., and Amos, B. (Joint first author)
41st International Conference on Machine Learning (ICML 2025)
Algorithms for mean-field variational inference via polyhedral optimization in the Wasserstein space
Jiang, Y., Chewi, S., and Pooladian, A-A. (Supervisory role)
Foundations of Computational Mathematics (2025+)
Entropic estimation of optimal transport maps
Pooladian, A-A., and Niles-Weed, J.
Best paper award at OTML NeurIPS workshop; 2021