Uniform almost flatness in finitely generated soluble groups

Abstract: We show that a finitely generated soluble group is virtually nilpotent if and only if the diameter of its finite coset spaces admits a uniform polynomial lower bound in terms of their size. We obtain the same conclusion for certain finitely generated abelian-by-cyclic groups under the weaker assumption that the diameters of their finite quotients are uniformly bounded below by a polynomial in their size. This extends the previous work of the author with Tointon.

• The paper proves that a finitely generated soluble group admits uniformly almost flat coset spaces if and only if it is virtually nilpotent.
• It employs a reduction to abelian-by-abelian groups combined with cohomological and number-theoretic techniques to analyze finite quotients.
• The study links coarse geometric properties with algebraic structure through uniform lower bounds on coset space diameters in finite quotients.

Uniform Almost Flatness in Finitely Generated Soluble Groups

Introduction and Context

The paper "Uniform almost flatness in finitely generated soluble groups" (2604.16866) provides a definitive answer to a conjecture relating uniform geometric properties—specifically, uniform lower bounds on diameters of finite coset spaces—to the algebraic structure of finitely generated soluble groups. Building on Gromov's polynomial growth theorem and earlier generalizations by Wolf, Milnor, and Tointon–Guo, this work characterizes virtual nilpotency purely in terms of "uniform almost flatness" conditions applied to coset spaces and finite quotients.

The main contribution is to prove that for any finitely generated soluble group, the existence of a polynomial uniform lower bound on the diameters of its finite coset spaces is equivalent to the group being virtually nilpotent. This equivalence sharpens the understanding of the interplay between algebraic and coarse geometric (metric) group properties in the solvable regime.

Preliminaries: Uniform Almost Flatness and Residual Finiteness

For a finitely generated group $G$ with filtration by finite-index (possibly normal) subgroups, two notions are key:

• Uniformly $\alpha$-almost flat coset spaces (u.c.$(\alpha)$): There is $\varepsilon > 0$ such that any finite-index subgroup $H < G$ satisfies

$$\operatorname{diam}_S(G/H) \geq \varepsilon [G : H]^\alpha$$

for a generating set $S$ and some $\alpha \in (0,1]$.

• Uniformly $\alpha$-almost flat quotients (u.q.$(\alpha)$): The same as above, but $\alpha$0 is restricted to finite-index normal subgroups.

It is shown that these properties are independent of the generating set and invariant under passage to finite-index subgroups and quotients. For groups of polynomial growth (degree $\alpha$1), u.c.$\alpha$2 holds by volume considerations.

The interaction between algebraic finiteness properties and the structure of subgroups and coset spaces is central; in particular, whether a non-polycyclic soluble group can maintain such uniform lower bounds relates to the classical Milnor–Wolf dichotomy, but here with uniformity replacing asymptotic growth.

Main Theorem and Structure of the Proof

The central result is:

Theorem: A finitely generated soluble group admits uniformly almost flat coset spaces if and only if it is virtually nilpotent.

The proof structure is intricate, involving several key reductions and delicate algebraic and number-theoretic arguments:

1. Reduction to Abelian-by-Abelian Groups: The argument reduces the general soluble case to groups that are abelian-by-abelian with a split extension using cohomological techniques and results on group structure, notably from Robinson–Wilson and Hall.
2. Finite Subgroup Rank Case: For groups with finite subgroup rank (all finitely generated subgroups require a bounded number of generators), the analysis uses tools from linear algebra (modeling group automorphisms by matrices over $\alpha$3) and number theory (control over the orders of eigenvalues and their reductions modulo large primes). If any automorphism in such an extension has "exotic" eigenvalues, i.e., those not roots of unity, it leads to construction of finite quotients with diameters violating the required lower bounds, using splitting primes and uniformity of associated coset space geometry.
3. Infinite Subgroup Rank Case: Here, Hall's results are used to guarantee the existence of infinite $\alpha$4-torsion in quotients. Explicit coset representatives are constructed whose small diameter (in terms of word metrics) compared to the index precludes any uniform polynomial lower bound, using arguments about bounded generation and width in abelian-by-polycyclic groups.
4. Obstructions via Classical Examples: The Baumslag-Solitar groups $\alpha$5 for $\alpha$6 and abelian wreath products $\alpha$7 (with $\alpha$8 finitely generated, nontrivial abelian) are shown not to have the u.q. property, via explicit construction of finite quotients whose diameter-to-size ratio violates any polynomial bound.

Each branch proceeds by carefully analyzing the structure of finite quotients and coset spaces, making use of the Breuillard-Green-Tao structure theory for approximate groups, modular representation theory, and uniformization of combinatorial group actions.

Key Technical Highlights

• Matrix and Module Structures: The abelian kernel in split extensions is realized as a rational $\alpha$9-vector space, and automorphisms by rational matrices. Finiteness and growth properties are encoded in whether these matrices are virtually unipotent or admit nontrivial roots of unity as eigenvalues.
• Number Theory of Reduction Modulo Primes: The proof leverages the existence of primes where characteristic polynomials split completely, and the divergence of multiplicative orders of eigenvalues over finite fields is crucial for constructing quotients of large order but small diameter.
• Use of Nilpotency Indices: The arguments for the negative direction (failure of uniform almost flatness in non-nilpotent cases) critically use the impossibility of uniformly controlling the index of nilpotence or the derived length in large finite quotients, as guided by Dixon's results on solvable permutation groups.

Consequences and Implications

Algebraic Consequences

The equivalence between the u.c. property and virtual nilpotency provides a new geometric characterization for virtually nilpotent finitely generated soluble groups within the broad class of residually finite groups. This places the polynomial growth dichotomy and the uniform diameter property within a single unifying framework.

Moreover, the methods highlight structural dichotomies (finite vs. infinite subgroup rank) that can be detected via geometric invariants defined purely via word metrics on finite quotients, linking properties of group rings, module generation, and filtration to large scale metric geometry.

Geometric Group Theory and Further Directions

The results reinforce the connection between the large-scale geometry of Cayley graphs and deep algebraic properties, and support conjectures (such as those by Tointon and collaborators) guessing that geometric regularity at the level of all finite quotients characterizes algebraic regularity (virtual nilpotency) in large classes.

Future developments may seek to extend these uniformity principles to larger classes: general residually finite groups or non-soluble settings, potentially involving right-angled Artin groups, branch groups, or groups with more intricate subgroup growth profiles. The technical machinery deployed here—intertwining number theory, algebraic group structure, and geometric analysis—serves as a potent template for further advancements in understanding rigidity phenomena in geometric group theory.

Conclusion

This work provides a precise and robust criterion for virtual nilpotency among finitely generated soluble groups, expressed through uniform lower bounds on coset space diameters of finite quotients. The arguments are comprehensive, utilizing reductions to well-understood algebraic cases, advanced number-theoretic input, and structural group theory. The implications are both conceptual—connecting metric geometry to group structure in a quantitative fashion—and technical, suggesting new avenues for the study of residual properties, subgroup growth, and large-scale geometry in discrete groups.