PreprintarXiv licencearXiv
Cutoff for random walks on nilpotent groups
Abstract
We study continuous-time random walks on finite nilpotent groups, allowing non-symmetric and nonuniform increment measures. Under a simple degree condition, sharp up to an absolute constant factor, we prove that total variation cutoff occurs if and only if the relaxation time is of smaller order than the mixing time. We establish a quantitative reduction to the Abelianization that preserves relaxation times and compares mixing profiles. Combining this reduction with a Gaussian approximation and the third-named author's cutoff criterion for diffusions yields the cutoff characterization.
Related
- Conditioned random walks on linear groups I: construction of the target harmonic measure
- Cutoff for random Cayley graphs of finite groups
- Stopping Rules and the Poisson Boundary of Random Walks on Countable Groups
- Large deviations for random walks generated by countable groups: a unified approach
- A nilpotent quotient algorithm for certain infinitely presented groups and its applications
BibTeX
@misc{hermon2026cutoff,
title = {{Cutoff for random walks on nilpotent groups}},
author = {Jonathan Hermon and Xiangying Huang and Justin Salez},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.35561},
eprint = {2610.35561},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.35561},
}
RIS
TY - UNPB
TI - Cutoff for random walks on nilpotent groups
AU - Jonathan Hermon
AU - Xiangying Huang
AU - Justin Salez
PY - 2026
DA - 2026/10/26
PB - arXiv
DO - 10.48550/arxiv.2610.35561
UR - https://arxiv.org/abs/2610.35561
AB - We study continuous-time random walks on finite nilpotent groups, allowing non-symmetric and nonuniform increment measures. Under a simple degree condition, sharp up to an absolute constant factor, we prove that total variation cutoff occurs if and only if the relaxation time is of smaller order than the mixing time. We establish a quantitative reduction to the Abelianization that preserves relaxation times and compares mixing profiles. Combining this reduction with a Gaussian approximation and the third-named author's cutoff criterion for diffusions yields the cutoff characterization.
ER -