Skip to main content
9 events
when toggle format what by license comment
Aug 14, 2019 at 18:21 comment added Jim Belk Note: In my previous comment, "path of length two as an isometrically embedded subgraph" should be "path of length two as an induced subgraph".
Aug 14, 2019 at 13:02 vote accept worldreporter
Aug 14, 2019 at 13:02 comment added worldreporter Thank you very much!
Aug 14, 2019 at 10:49 comment added Jim Belk (continued) Indeed, it is not hard to show that any graph $\Gamma$ which does not have a path of length two as an isometrically embedded subgraph must be a disjoint union of complete graphs. It follows that a Gromov hyperbolic, right-angled Coxeter group $W_\Gamma$ has the property you want if and only if it is a free product of finite groups.
Aug 14, 2019 at 10:41 comment added Jim Belk The group $G$ given at the end above corresponds to the right-angled Coxeter group $W_\Gamma$ for which $\Gamma$ is a path of length two. More generally, if $\Gamma$ is any graph which has a path of length two as an induced subgraph, then $W_\Gamma$ will have $G$ as an isometrically embedded subgroup, and the same phenomenon will occur. For example, $\mathbb{Z}_2* G$ is hyperbolic and irreducible (not a direct product -- I'm assuming that's what you mean by "irreducible") , and its Cayley graph has the Cayley graph of $G$ as an isometrically embedded subgraph.
Aug 14, 2019 at 6:18 comment added worldreporter Thank you for your great answer! Is it possible that at least in the right-angled, irreducible case my question has a positive answer?
Aug 13, 2019 at 21:41 history edited Jim Belk CC BY-SA 4.0
deleted 12 characters in body
Aug 13, 2019 at 20:40 history edited Jim Belk CC BY-SA 4.0
added 1022 characters in body
Aug 13, 2019 at 20:27 history answered Jim Belk CC BY-SA 4.0