Further and stronger analogy between sampling and optimization: Langevin Monte Carlo and gradient descent Leave a comment CREST Working Papers Series No. 2017-21 by Arnak Dalalyan