1: \begin{abstract}

2: The logarithmic divergence is an extension of the Bregman divergence

3: motivated by optimal transport and a generalized convex duality, and satisfies

4: many remarkable properties. Using the geometry induced by the logarithmic

5: divergence, we introduce a generalization of continuous time mirror descent that

6: we term the conformal mirror descent. We derive its dynamics under a generalized

7: mirror map, and show that it is a time change of a corresponding Hessian

8: gradient flow. We also prove convergence results in continuous time. We apply

9: the conformal mirror descent to online estimation of a generalized exponential

10: family, and construct a family of gradient flows on the unit simplex via the

11: Dirichlet optimal transport problem.

12: \end{abstract}

13: