1: \begin{abstract}

2: {Inexact Markov Chain Monte Carlo methods rely on Markov chains that 

3: do not exactly preserve the target distribution. Examples include the unadjusted Langevin algorithm (ULA) and unadjusted Hamiltonian Monte Carlo (uHMC).

4:  This paper establishes bounds on Wasserstein distances between the

5: invariant probability measures 

6: of inexact MCMC methods and their target distributions with a focus on understanding the precise dependence of this asymptotic bias on both

7: dimension and discretization step size.

8: Assuming Wasserstein bounds on the convergence to equilibrium of 

9: either the exact or the approximate dynamics, we show that for both

10: ULA and uHMC, the asymptotic bias depends on key quantities related to the target distribution or the stationary probability measure of the scheme. As a corollary, we conclude that for models with a limited amount of interactions such as mean-field models, finite range graphical models, and

11: perturbations thereof, the asymptotic bias has a similar 

12: dependence on the step size and the dimension as for product measures.

13: %several important classes of models

14: %(including for example mean-field models, finite range graphical models and

15: %perturbations of Gaussian measures). 

16: % In particular, if dimension-free bounds

17: % on convergence to equilibrium hold in a standard Wasserstein distance,

18: % then for such models, the resulting bounds for the asymptotic bias are of order $O(d^{1/2}\gamma )$ for ULA and $O(d^{1/2}\gamma^2 )$ for uHMC.

19: % On the other hand,

20: % for more general models, the dimension dependence of the asymptotic bias may be worse than in

21: % the product case even  if the exact dynamics has dimension-free mixing properties.

22: }

23: \end{abstract}

24: