摘要:
The framework of spectral independence and the stochastic localization technique are two recent techniques for proving mixing bounds for Markov chains. The framework of localization schemes unifies and extends the two techniques and provides tools for deriving mixing bounds for Markov chains through the analysis of the corresponding localization process.
In this talk, we will briefly introduce the framework of localization schemes and its applications based on the work of Yuansi Chen and Ronen Eldan[2022].