Mean-field Gibbs Measures: Sharp and Optimal Propagation of Chaos
发布时间：2023年09月05日
浏览次数：1876
发布者: He Liu
主讲人： Rishabh Gvalani (the Max Planck Institute in Leipzig)
活动时间： 从 2023-09-06 16:00 到 17:00
场地： 线上
We study the invariant Gibbs measure of mean-field interacting diffusions and prove optimal rates of convergence to its thermodynamic limit in the full subcritical regime of temperatures $T>T_c$. Our proof relies on a non-asymptotic Sanov-type upper bound which is of independent interest itself. We also apply these techniques to prove sharp exponential concentration inequalities for i.i.d empirical measures in negative Sobolev norms. This is joint work with Matías G. Delgadino.
Oline talk via Zoom
https://mpimis.zoom.us/j/5271342110?pwd=SE5zN1kwYmcrazRjSEQ2Qnl0YWJYdz09
meeting number (for use in app): 5271342110
participant passcode: 7342224
