Nic Fishman

Papers

Total Citations

H-Index

About

Nic Fishman is a rising researcher at the forefront of generative modeling, specializing in the development of diffusion models for complex, constrained domains. Their work tackles a critical bottleneck in modern AI: extending state-of-the-art generative algorithms beyond standard image spaces to Riemannian manifolds—geometrically structured data prevalent in the natural sciences. Fishman’s major contributions include pioneering the first diffusion models tailored for such constrained environments, a breakthrough that unlocks powerful generative capabilities for problems in physics, biology, and chemistry. Their 2023 paper, “Diffusion Models for Constrained Domains,” lays the theoretical foundation for this domain, while the follow-up, “Metropolis Sampling for Constrained Diffusion Models,” introduces a novel sampling algorithm that dramatically improves the efficiency and accuracy of these models. Though early in their career, Fishman’s work has already garnered significant attention, with these key papers accumulating 8 citations and establishing a new research trajectory. By bridging the gap between high-performance generative AI and the structured data of the natural world, Nic Fishman is poised to become a leading voice in the next wave of scientific machine learning.