1: \begin{abstract}

2: We consider a discrete-time system of $n$ coupled random vectors,

3: a.k.a.~interacting particles.  The dynamics involve a vanishing step size,

4: some random centered perturbations, and a mean vector field which induces the

5: coupling between the particles.  We study the doubly asymptotic regime where

6: both the number of iterations and the number $n$ of particles tend to infinity,

7: without any constraint on the relative rates of convergence of these two

8: parameters.  We establish that the empirical measure of the interpolated

9: trajectories of the particles converges in probability, in an ergodic sense, to

10: the set of recurrent Mc-Kean-Vlasov distributions.  We also consider the

11: pointwise convergence of the empirical measures of the particles.  A first

12: application example is the granular media equation, where the particles are

13: shown to converge to a critical point of the Helmholtz energy.  A second

14: example is the convergence of stochastic gradient descent to the global

15: minimizer of the risk, in a wide two-layer neural networks using random

16: features.  

17: \end{abstract}

18: