Continuity of asymptotic entropy on wreath products

Eduardo Silva · 2025

Read Continuity of asymptotic entropy on wreath products by Eduardo Silva (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-01712.txt.

Introduction

Eduardo Silva proves smooth drift of long-run walk spread on wreath product groups in 2025. Such groups stack one group on a base group, and walk spread is read by a rate called asymptotic entropy. The base group must hold a fast-growth fineness and a tame center class.

One step shows the chance a walk never comes home shifts smooth with the step law. A wide step class on known groups then falls under the same smooth law. The gain covers line groups and groups that act on squared-space forms as well.

Catalog record

CREATOREduardo Silva
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE146 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-01712

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Continuity of asymptotic entropy on wreath products Eduardo Silva Abstract . We prove the continuity of asymptotic entropy as a function of the step distribution for non-degenerate probability measures with finite entropy on wreath products A ≀ B = ⨁ B A ⋊ B A\wr B=\bigoplus_{B}A\rtimes B , where A A is any countable group and B B is a countable hyper-FC-central group that contains a finitely generated subgroup of at least cubic growth. As one step in proving the above, we show that on any countable group G G the probability that the μ \mu -random walk on G G never returns to the identity is continuous in μ \mu , for measures μ \mu such that the semigroup generated by the support of μ \mu contains a finitely generated subgroup of at least cubic growth. Finally, we show that among random walks on a group G G that admit a separable completely metrizable space X X as a model for their Poisson boundary, the weak continuity of the associated harmonic measures on X X implies the continuity of the asymptotic entropy. This result recovers the continuity of asymptotic entropy on known cases, such as Gromov hyperbolic groups and acylindrically hyperbolic groups, and extends it to new classes of groups, including linear groups and groups acting on CAT ⁡ ( 0 ) \mathrm{CAT}(0) spaces. 2020 Mathematics Subject Classification 20F69, 60B15, 60J50, 20F65, 43A05 1. Introduction Asymptotic entropy is a fundamental quantity in understanding the long-term behavior of random walks on countable infinite groups. For a probability measure \mu on a group G , the Shannon entropy of \mu is defined as H(\mu)\coloneqq-\sum_{g\in G}\mu(g)\log(\mu(g)) . Definition 1.1 . The asymptotic entropy of a probability measure \mu on a countable group G is defined as h(\mu)\coloneqq\lim_{n\to\infty}\frac{H(\mu^{*n})}{n}. This notion was introduced by Avez [ Ave72 ] , who proved that if \mu is finitely supported and h(\mu)=0 , then the random walk (G,\mu) has the Liouville property , meaning that every bounded \mu -harmonic function f:G\to\mathbb{R} is constant on the cosets of \langle\mathrm{supp}(\mu)\rangle . We recall that a function f:G\to\mathbb{R} is called \mu -harmonic if f(g)=\sum_{h\in G}f(gh)\mu(h) , for all g\in G . The entropy criterion of Derriennic [ Der80 ] and Kaimanovich-Vershik [ KV83 , Theorem 1.1] states that if H(\mu)<\infty , then h(\mu)>0 if and only (G,\mu) does not have the Liouville property. Recall that \mu is called non-degenerate if G=\langle\mathrm{supp}(\mu)\rangle_{+} (the semigroup generated by \mathrm{supp}(\mu) ). If G is a non-amenable group, then any non-degenerate probability measure \mu on G does not have the Liouville property [ Aze70 , Proposition II.1] [ Fur73 ] . Hence, if in addition H(\mu)<\infty , then h(\mu)>0 . Reciprocally, every amenable group admits a non-degenerate symmetric probability measure \mu with the Liouville property [ Ros81 , Theorem 1.10] [ KV83 , Theorem 4.4] . In general, one cannot guarantee in this statement that H(\mu)<\infty . Indeed, there are amenable groups on which every non-degenerate probability measure \mu with H(\mu)<\infty does not have the Liouville property [ Ers04 , Theorem 3.1] . The value h(\mu) determines that, asymptotically, the n -th instant of the \mu -random walk belongs to a subset of G of size roughly \exp(h(\mu)n) ; see Remark 2.4 for a precise statement. For random walks with a finite first moment on a finitely generated group, another related quantity is the asymptotic drift \ell(\mu)\coloneqq\lim_{n\to\infty}\mathbb{E}[|w_{n}|]/n of the random walk \{w_{n}\}_{n\geq 0} with respect to a word metric |\cdot| . In many classes of groups, the ratio h(\mu)/\ell(\mu) coincides with the Hausdorff dimension of the hitting measure of the \mu -random walk on a geometric boundary of the group G . This is the case for free groups [ Kai98 , Theorem 2.1.3] and hyperbolic groups [ LP08 , Theorem 3.1] ; see [ Tan19 , DY23 ] for further families of groups where this equality holds. We also refer the reader to [ Ver00 ] for a more extensive explanation on the connection between the asymptotic entropy of random walks and other dynamical and algebraic properties of the group. The question of whether the function \mu\mapsto h(\mu) on a given group G is continuous goes back to [ KV83 , Section 6, Remark 3] ; see also [ Ers04 , Section 6, Item 5] for an explicit formulation of this problem, and [ Ers11a ] , where the first positive results on continuity of asymptotic entropy are proved. One motivation for this question is that there are multiple situations exhibiting discontinuity in asymptotic entropy within some classes of (amenable) groups. The first examples of such discontinuity were given by Kaimanovich [ Kai83 , Theorems 3.1 and 4.1] , who showed that certain locally finite groups – such as the group of finitely supported permutations of a countable set and the wreath product \mathbb{Z}/2\mathbb{Z}\wr(\bigoplus_{\mathbb{N}}\mathbb{Z}/2\mathbb{Z}) – admit infinitely supported probability measures with finite entropy, and which do not have the Liouville property. By the entropy criterion, such measures have positive asymptotic entropy, which cannot be approximated by the asymptotic entropy of finitely supported measures. The above examples involve infinitely supported measures, but discontinuity can also occur for probability measures supported on a fixed finite symmetric subset of a group. Indeed, such an example is given in [ Ers11a , Lemma 4] on the group \mathbb{Z}/2\mathbb{Z}\wr D_{\infty} , where D_{\infty} is the infinite dihedral group. 1.1. Continuity of entropy on wreath products Our first main result shows that asymptotic entropy is continuous on wreath products A\wr B\coloneqq\bigoplus_{B}A\rtimes B where the base group B is a hyper-FC-central group with sufficiently fast growth. Recall that a group G is called hyper-FC-central if G has no non-trivial quotient on which any non-trivial element has an infinite conjugacy class. Recall also that a finitely generated group G is said to have at least cubic growth if there is a finite symmetric generating set S of G and a constant C>0 such that |(S\cup\{e_{G}\})^{n}|\geq Cn^{3} for all n\geq 1 . Theorem 1.2 . Let A be a countable group and let B be a countable hyper-FC-central group that contains a finitely generated subgroup of at least cubic growth. Let \mu be a non-degenerate probability measure \mu on A\wr B with H(\mu)<\infty . Consider a sequence \{\mu_{k}\}_{k\geq 1} of probability measures on A\wr B with H(\mu_{k})<\infty for all k\geq 1 , such that \lim_{k\to\infty}\mu_{k}(g)=\mu(g) for every g\in A\wr B and \lim_{k\to\infty}H(\mu_{k})=H(\mu) . Then, we have \lim_{k\to\infty}h(\mu_{k})=h(\mu) . In particular, Theorem 1.2 applies when B=\mathbb{Z}^{d} , for d\geq 3 . This is the first known result that establishes the continuity of asymptotic entropy for amenable groups that are not hyper-FC-central, and which works for probability measures with an infinite support. Additionally, it is the first result to address arbitrary sequences of finitely supported probability measures on an amenable group. The only prior result in this direction is [ Ers11a , Theorem 5] , which shows the continuity of asymptotic entropy for A\wr\mathbb{Z} with A finite, under the assumption that the supports of all probability measures \mu , \mu_{k} , k\geq 1 , are contained in a fixed finite subset of A\wr\mathbb{Z} . Another feature of Theorem 1.2 is that it holds for all non-degenerate probability measures with finite entropy. For all such measures,

Read the complete text