Sophia Tang

Expanding Flow Maps

Sophia Tang, Pranam Chatterjee

A2D2: Fine-Tuning Any-Length Discrete Diffusion for Adaptive Decoding

Sophia Tang, Yuchen Zhu, Molei Tao, et al.

Foundations of Schrödinger Bridges for Generative Modeling

Sophia Tang

This paper explains Schrödinger bridges, a mathematical framework that unifies popular AI generative models like diffusion models and flow matching. The core idea is finding the optimal way to transform a simple distribution into a complex one while minimizing entropy deviation, which the paper develops from first principles using tools from optimal transport and stochastic control. The work provides both theoretical foundations and practical methods for building these bridges, connecting abstract mathematics to modern generative AI techniques.