English: The contours in this video represent level sets of a continuous probability distribution and we consider the task of summarising this distribution with a discrete set of states selected from its domain. In particular, we suppose that the density function is known up to proportionality, a setting where Markov chain Monte Carlo (MCMC) methods are widely used. In the first part of this video a Markov chain explores the high-probability region, with the sample path shown in black. In the second part of the video an algorithm, called Stein Thinning, is applied to select a subset of states from the sample path, such that together these states provide an accurate approximation of the continuous probability distribution. See Riabiz et al, "Optimal Thinning of MCMC Output", in the Journal of the Royal Statistical Society Series B.