Shakir Mohamed

Papers

Total Citations

H-Index

About

Shakir Mohamed is a leading researcher at the intersection of machine learning, generative models, and creative AI. His work fundamentally advances density estimation and probabilistic modeling, most notably through his pioneering contributions to normalizing flows. His 2016 paper, "Normalizing Flows on Riemannian Manifolds" (46 citations), introduced a powerful framework for density estimation on complex geometric spaces, with applications spanning fluid mechanics, optics, and computational biology. This work helped establish normalizing flows as a cornerstone of modern generative modeling. More recently, Mohamed has pushed the boundaries of AI’s creative potential, exemplified by his 2024 study "A Robot Walks into a Bar" (33 citations). This unique collaboration with professional comedians at the Edinburgh Festival Fringe explored how large language models can serve as creativity support tools for comedy, evaluating their humor alignment with human performers. By bridging rigorous mathematical theory with human-centered applications, Mohamed demonstrates a rare ability to advance both fundamental science and socially impactful technology. His research continues to inspire students and researchers working at the frontiers of probabilistic machine learning and AI-driven creativity.