1: \begin{abstract}

2: Langevin MCMC gradient optimization is a class of increasingly popular methods for estimating a posterior distribution.

3: This paper addresses the algorithm as applied in a decentralized setting, wherein data is distributed across

4: a network of agents which act to cooperatively solve the problem using peer-to-peer gossip communication.

5: We show, theoretically, results in 1) the time-complexity to $\epsilon$-consensus for the continuous

6: time stochastic differential equation, 2) convergence rate in $L^2$ norm to consensus for the discrete implementation

7: as defined by the Euler-Maruyama discretization and 3) convergence rate in the Wasserstein metric to the optimal stationary

8: distribution for the discretized dynamics.

9: \end{abstract}

10: