Reducing the conjugacy problem for relatively hyperbolic automorphisms to peripheral components

From MaRDI portal





Let \(G\) be a residually finite torsion-free hyperbolic group and \(\alpha\) an outer automorphism of \(G\). The dynamics of \(\alpha\) distinguish polynomially growing elements, which form maximal subgroups by a result of Levitt and Lustig. For each such maximal subgroup \(A\), there exists minimal \(k_A > 0\) and \(g_A \in G\) such that \(A\) is invariant under \(\beta = \mathrm{ad}_{g_A} \circ \alpha^{k_A}\), yielding the sub-mapping-torus \(A \rtimes_\beta \mathbb{Z}\).\N\NTheorem A reduces the conjugacy problem in \(\mathrm{Out}(G)\) to algorithmic problems for these polynomial growth sub-mapping-tori and their JSJ edge groups. If these sub-mapping-tori satisfy four properties (hereditarily algorithmically tractable, effective congruences separating torsion, solvable fiber-and-orientation-preserving isomorphy problem, solvable mixed Whitehead problem for edge group tuples), then the conjugacy problem is solvable. The proof uses relative hyperbolicity of \(G \rtimes_\alpha \mathbb{Z}\) and two types of Dehn fillings: co-finite fillings for JSJ twisting and co-virtually-cyclic fillings preserving fibered structure.\N\NApplications include outer automorphisms of \(F_n\) that are pointwise inner on polynomially growing elements and similar results for one-ended residually finite torsion-free hyperbolic groups. Proposition B verifies the properties when sub-mapping-tori are direct products of free groups with \(\mathbb{Z}\).



Cites work









This page was built for publication: Reducing the conjugacy problem for relatively hyperbolic automorphisms to peripheral components

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6914774)