arXiv2025
These notes expand upon our lectures on {\em profinite rigidity} at the international colloquium on randomness, geometry and dynamics, organised by TIFR Mumbai at IISER Pune in January 2024. We are interested in the extent to which groups that arise in hyperbolic geometry and 3-manifold topology are determined by their finite quotients. The main theme of these notes is the radical extent to which rigidity is lost when one passes from consideration of groups with hyperbolic features to consideration of their direct products. We describe a general method for producing infinite sequences of {\em{Grothendieck pairs,}} i.e.~embeddings $P_i\hookrightarrow G\times G$ inducing isomorphisms of profinite completions, with $G$ fixed and $P_i$ finitely generated. In order to apply this method, one needs $G$ to map onto a subgroup of finite index in the commutator subgroup of a group $Γ$ with $H_2(Γ,\mathbb{Z})=0$, and $Γ$ should be relatively hyperbolic. By exploiting the flexibility of the construction, we explain how, under the same hypotheses on $G$, one can construct {\em uncountable families} of pairwise non-isomorphic subgroups $P_λ$ such that $P_λ\hookrightarrow G\times G$ induces an isomorphism of profinite completions. Examples of groups $G$ satisfying these conditions include the fundamental group of the Weeks manifold and the fundamental group of the 4-fold branch cover of the figure-8 knot complement. Both of these examples are profinitely rigid in the absolute sense and in each case Grothendieck pairs account entirely for the loss of profinite rigidity for $G\times G$: if $H$ is a finitely generated group whose profinite completion is isomorphic to that of $G\times G$, then there is an embedding $H\hookrightarrow G\times G$ that is a Grothendieck pair.