Lower Bounds on Metropolized Sampling Methods for Well-Conditioned Distributions

Authors: Yin Tat Lee, Ruoqi Shen, Kevin Tian

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Theoretical We give lower bounds on the performance of two of the most popular sampling methods in practice, the Metropolis-adjusted Langevin algorithm (MALA) and multi-step Hamiltonian Monte Carlo (HMC) with a leapfrog integrator, when applied to well-conditioned distributions. Our main result is a nearly-tight lower bound of eΩ(κd) on the mixing time of MALA from an exponentially warm start, matching a line of algorithmic results [DCWY18, CDWY19, LST20a] up to logarithmic factors and answering an open question of [CLA+20].
Researcher Affiliation Collaboration Yin Tat Lee University of Washington and Microsoft Research yintat@uw.edu Ruoqi Shen University of Washington shenr3@cs.washington.edu Kevin Tian Stanford University kjtian@stanford.edu
Pseudocode No The paper describes algorithms (MALA, HMC) using mathematical formulations and textual descriptions but does not provide structured pseudocode blocks or algorithm listings.
Open Source Code No The paper does not provide any concrete access information (e.g., repository link or explicit statement of code release) for the methodology described.
Open Datasets No This is a theoretical paper and does not involve empirical studies with datasets, thus no information on public dataset access is relevant or provided.
Dataset Splits No This is a theoretical paper and does not involve empirical studies with dataset splits.
Hardware Specification No This is a theoretical paper and does not describe any experimental setup involving specific hardware.
Software Dependencies No The paper does not provide specific ancillary software details with version numbers used for its research.
Experiment Setup No This is a theoretical paper and does not describe any experimental setup details such as hyperparameters or training configurations.