Uniform-in-$N$ log-Sobolev inequality for the mean-field Langevin dynamics with convex energy
Sinho Chewi, Atsushi Nitanda, Matthew S. Zhang

TL;DR
This paper proves a uniform log-Sobolev inequality for mean-field Langevin dynamics, showing the constant does not depend on the number of particles, using a Lipschitz transport map and reverse heat flow techniques.
Contribution
It introduces a novel uniform-in-$N$ log-Sobolev inequality for mean-field Langevin dynamics with convex energy functions.
Findings
Established a uniform log-Sobolev inequality independent of particle number $N$
Developed a Lipschitz transport map via reverse heat flow
Provided new tools for analyzing mean-field stochastic processes
Abstract
We establish a log-Sobolev inequality for the stationary distribution of mean-field Langevin dynamics with a constant that is independent of the number of particles . Our proof proceeds by establishing the existence of a Lipschitz transport map from the standard Gaussian measure via the reverse heat flow of Kim and Milman.
Peer Reviews
No public reviews on file for this paper yet. If you reviewed it on a platform where reviews are public (OpenReview, ICLR, NeurIPS, ICML), you can paste yours below so the community can read it here.
Videos
No videos yet. Explain this paper in a talk, walkthrough, or lecture? Add one.
Taxonomy
TopicsMarkov Chains and Monte Carlo Methods · Random Matrices and Applications · Statistical Methods and Inference
