Gradient estimates for the Schrodinger potentials: convergence to the Brenier map and quantitative stability

We show convergence of the gradients of the Schrodinger potentials to the Brenier map in the small-time limit under general assumptions on the marginals, which allow for unbounded densities and supports. Furthermore, we provide novel quantitative stability estimates for the optimal values and optimal couplings for the Schrodinger problem (SP), that we express in terms of a negative order weighted homogeneous Sobolev norm. The latter encodes the linearized behavior of the 2-Wasserstein distance between the marginals. The proofs of both results highlight for the first time the relevance of gradient bounds for Schrodinger potentials, that we establish here in full generality, in the analysis of the short-time behavior of Schrodinger bridges. Finally, we discuss how our results translate into the framework of quadratic Entropic Optimal Transport, that is a version of SP more suitable for applications in machine learning and data science.

Automated summary

In this paper, we prove the convergence of the gradients of Schrodinger potentials to the Brenier map in the small-time limit, under general assumptions on the marginals. We also provide novel quantitative stability estimates for the optimal values and optimal couplings for the Schrodinger problem, expressed in terms of a negative order weighted homogeneous Sobolev norm. Our results highlight the relevance of gradient bounds for Schrodinger potentials and their application in the framework of quadratic Entropic Optimal Transport.