Rigid Relative Automorphism Groups

• Rigid relative automorphism groups are families of automorphism structures constrained by fixed peripheral data and rigidity conditions, resulting in controlled group extensions.
• They arise in diverse settings like affine varieties, character varieties, RAAGs, and linking systems, where structured subdata yield explicit automorphism classifications.
• This framework replaces complex automorphism groups with manageable forms such as torus actions, finite extensions, and split sequences, enabling inductive analysis.

A rigid relative automorphism group is not a single universally standardized object, but rather a family of closely related constructions in which automorphisms are constrained by both a rigidity condition and a relative condition. In the cited literature, the relative condition typically requires automorphisms to preserve or fix distinguished data such as boundary monodromies, special subgroups, a defining function, clique subalgebras, or a Sylow/fusion system, while rigidity means either the absence of nontrivial additive symmetries, the smallness of automorphisms acting trivially on that data, or the persistence of prescribed local structure. The resulting groups are often highly structured: finite extensions of maximal tori, split extensions of mapping class groups, elementary abelian $2$-groups, or recursively describable relative outer automorphism groups (Arzhantsev et al., 2016, Kim, 13 Aug 2025, Glauberman et al., 2019).

1. Terminological scope and recurrent structure

The expression appears in several non-equivalent settings. What remains common is that the ambient automorphism group is filtered through fixed peripheral data, and rigidity sharply limits what survives. A frequent misconception is that “rigid” means “trivial automorphism group.” In the literature summarized here, rigidity often still permits torus actions, finite permutation groups, inner automorphisms, or explicitly controlled local symmetries.

Setting	Relative condition	Structural conclusion
Rigid affine varieties	Preserve torus grading or a function $f$ up to scalar	Unique maximal torus; often a finite extension of it (Arzhantsev et al., 2016, Borovik et al., 2020)
$\mathrm{SL}_2$-character varieties	Fix traces of puncture monodromies	Split extension of $\Mod(\Sigma)$ by $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$ (Kim, 13 Aug 2025)
RAAGs	Preserve $\mathcal G$, act trivially on $\mathcal H$	Restriction kernels and images are again relative RAAG automorphism groups (Day et al., 2017)
Linking systems	Restrict to the identity on the Sylow object or fusion system	Trivial at odd primes; elementary abelian at $p=2$, with split extension over rigid inner automorphisms (Glauberman et al., 2019)
Graph product von Neumann algebras	Respect clique/intersection subalgebras	Automorphisms are generated by group, character, clique-local, and inner data (Chifan et al., 2022)

This pattern suggests that “rigid relative automorphism group” is best understood as a methodological theme rather than a single definition: one first singles out canonical substructures, then proves that automorphisms compatible with them admit a small or explicit description.

2. Algebraic-geometric rigidity: maximal tori, separable variables, and function-preserving automorphisms

For affine algebraic varieties, rigidity means the absence of nontrivial $\mathbb G_a$-actions. Equivalently, the coordinate ring admits no nonzero locally nilpotent derivations, and the Makar-Limanov invariant is maximal (Arzhantsev et al., 2016). A foundational result states that if $X$ is a rigid affine variety, then $f$0 contains a unique maximal torus $f$1, normal in $f$2. When the canonical grading on $f$3 defined by $f$4 is pointed, $f$5 is a finite extension of $f$6; in particular, the neutral component is toral and the residual symmetry is finite (Arzhantsev et al., 2016).

This general structure becomes explicit for hypersurfaces with separable variables. Such a hypersurface is defined by a polynomial $f$7 in which each variable appears in exactly one monomial. For a rigid hypersurface $f$8 of this type, the automorphism group is described as

$f$9

where $\mathrm{SL}_2$0 is the finite permutation group preserving the monomial pattern of $\mathrm{SL}_2$1, and $\mathrm{SL}_2$2 is the subgroup of the linear torus preserving the zero locus of $\mathrm{SL}_2$3 (Trushin, 2024). The paper shows, in particular, that $\mathrm{SL}_2$4 is a finite extension of the maximal torus. The argument proceeds by analyzing invariant subvarieties and variable blocks, and by excluding triangular or unipotent automorphisms because any such automorphism would yield a nontrivial $\mathrm{SL}_2$5-action, contradicting rigidity (Trushin, 2024).

A model case is the Fermat-type hypersurface $\mathrm{SL}_2$6 with $\mathrm{SL}_2$7, whose group of linear automorphisms is

$\mathrm{SL}_2$8

Here $\mathrm{SL}_2$9 permutes the variables, the finite cyclic factors arise from root-of-unity scalings, and $\Mod(\Sigma)$0 is the diagonal torus compatible with the defining equation (Trushin, 2024). This is a representative instance in which rigidity does not eliminate automorphisms, but compresses them into a torus-plus-finite form.

A more explicitly relative construction appears for $\Mod(\Sigma)$1-suspensions. Given a rigid variety $\Mod(\Sigma)$2, a regular function $\Mod(\Sigma)$3, and positive integers $\Mod(\Sigma)$4, one defines

$\Mod(\Sigma)$5

The relevant relative automorphism group is

$\Mod(\Sigma)$6

If $\Mod(\Sigma)$7, then $\Mod(\Sigma)$8 is rigid, and

$\Mod(\Sigma)$9

where $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$0 is the natural scaling torus; if $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$1, then $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$2 is a finite extension of $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$3 (Borovik et al., 2020). This relative viewpoint isolates the subgroup of automorphisms compatible with the distinguished function $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$4, and then lifts it to the suspended rigid variety.

3. Boundary-relative rigidity on character varieties

For a punctured surface $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$5 with fundamental group $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$6, the $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$7-character variety $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$8 is the GIT quotient of the representation variety by conjugation. Let $H^1(\Sigma,\partial\Sigma;\mathbb Z/2)$9 denote curves homotopic to the punctures. The relative automorphism group is defined by fixing the trace functions of these boundary monodromies: $\mathcal G$0 Equivalently, it is the automorphism group of the character algebra over the subalgebra generated by the boundary traces (Kim, 13 Aug 2025).

The central structural result is a split exact sequence

$\mathcal G$1

for $\mathcal G$2, with minor adjustments in the exceptional cases $\mathcal G$3 and $\mathcal G$4 (Kim, 13 Aug 2025). Thus the rigid relative automorphism group is a finite extension of the mapping class group by the finite group of central representations. In this formulation, rigidity means that any algebra automorphism fixing the puncture data is forced to arise from a mapping class, up to this finite central ambiguity.

The proof strategy uses the curve complex and measured lamination theory. The curve complex rigidity theorems of Ivanov, Korkmaz, and Luo are invoked to promote combinatorial control of simple closed curves to control of the character variety automorphisms, while embeddings of measured laminations into the Berkovich boundary show that enough geometric data is preserved to identify automorphisms with mapping classes and central sign changes (Kim, 13 Aug 2025). Concretely, the paper states that for a simple closed curve $\mathcal G$5, an automorphism sends $\mathcal G$6 to $\mathcal G$7, where $\mathcal G$8 is a mapping class and the sign comes from a central representation.

For surfaces with one or two boundary components, the paper also exploits an exceptional isomorphism between the $\mathcal G$9-character variety and a moduli space of points on the complex $\mathcal H$0-sphere. This yields a new description of the mapping class group of certain $\mathcal H$1 in terms of automorphisms of that moduli space (Kim, 13 Aug 2025). In this setting, the adjective “relative” is literal: the puncture monodromies are the fixed peripheral parameters.

4. Relative outer automorphism groups of right-angled Artin groups

For a RAAG $\mathcal H$2 with defining graph $\mathcal H$3, and families $\mathcal H$4 of special subgroups, the relative outer automorphism group is

$\mathcal H$5

the subgroup of $\mathcal H$6 consisting of outer automorphisms that preserve each subgroup in $\mathcal H$7 up to conjugacy and act trivially on each subgroup in $\mathcal H$8 (Day et al., 2017). This is one of the most explicit uses of the term “relative automorphism group”: the peripheral structure is built into the definition.

The finite-index subgroup $\mathcal H$9 is generated by the Laurence-type generators it contains: inversions, transvections, and extended partial conjugations (Day et al., 2017). The paper gives precise inclusion criteria. For example, an inversion $p=2$0 lies in the relative group exactly when $p=2$1 does not lie in any subgroup of $p=2$2; a transvection $p=2$3 lies in the relative group when the relative domination condition $p=2$4 holds; and an extended partial conjugation is allowed when its support is a union of the appropriate $p=2$5-components of $p=2$6 (Day et al., 2017). These criteria make the relative group algorithmically accessible in terms of the defining graph and the chosen peripheral families.

The main technical theorem concerns restriction homomorphisms. If $p=2$7 is preserved and $p=2$8 is saturated with respect to $p=2$9, there is an exact sequence

$\mathbb G_a$0

The decisive rigidity feature is that both kernel and image are again relative outer automorphism groups of RAAGs, but on simpler data (Day et al., 2017). This closure under restriction enables induction on the graph and explains the subnormal decomposition of $\mathbb G_a$1.

The resulting structure theorem gives a subnormal series in which each quotient is finite, free-abelian, $\mathbb G_a$2, or a Fouxe-Rabinovitch group, and from this the authors deduce that $\mathbb G_a$3 is type VF (Day et al., 2017). In this context, rigidity is not a claim of small cardinality; it is the self-similarity of the class of relative automorphism groups under restriction and decomposition.

5. Rigid automorphisms of linking systems and related finite-group data

In the theory of saturated fusion systems and centric linking systems, a rigid automorphism is one that restricts to the identity on the Sylow $\mathbb G_a$4-subgroup or, equivalently in the transporter-system formulation, on the distinguished Sylow object. The subgroup of rigid inner automorphisms is given by conjugation by elements of $\mathbb G_a$5, the center of the Sylow subgroup $\mathbb G_a$6 (Glauberman et al., 2019).

The main structural theorem states that for a linking locality or linking system at a prime $\mathbb G_a$7, the group of rigid outer automorphisms $\mathbb G_a$8 is abelian of exponent at most

$\mathbb G_a$9

Hence rigid outer automorphisms are trivial at odd primes and form an elementary abelian $X$0-group at $X$1. Moreover, the short exact sequence

$X$2

splits (Glauberman et al., 2019). The same work identifies $X$3 with the derived limit $X$4, providing a cohomological description of rigid outer automorphisms. Another theorem shows that if $X$5 is a finite group with $X$6, then any outer automorphism of $X$7 that restricts to the identity on the centric linking system is of $X$8-order modulo inner automorphisms (Glauberman et al., 2019).

A later classification determines when non-inner rigid automorphisms actually occur for centric linking systems of finite groups of Lie type in odd characteristic at the prime $X$9. For a finite quasisimple group $f$00 of Lie type over a field of odd characteristic, with $f$01 the associated centric linking system at $f$02, non-inner rigid automorphisms exist exactly for two families:

1. $f$03 with $f$04;
2. $f$05 with $f$06 and $f$07. In these cases, $f$08 (Villareal, 25 Sep 2025). The corollary for all known quasisimple saturated fusion systems at $f$09 includes corresponding alternating-group exceptions as well (Villareal, 25 Sep 2025). This classification sharpens the earlier structural theorem: the only possible failure of “all rigid automorphisms are inner” is now isolated in explicit families.

6. Other rigid-relative frameworks and comparative perspective

In operator algebras, rigidity of automorphisms appears for group von Neumann algebras $f$10 of certain graph product groups whose underlying graph is a cycle of cliques and whose vertex groups are wreath-like property $f$11 groups. Every automorphism is, up to inner conjugacy, of the form

$f$12

or equivalently $f$13, where $f$14 comes from a group automorphism and a character, while $f$15 is a clique-local automorphism implemented by unitaries in clique or intersection subalgebras (Chifan et al., 2022). Under additional hypotheses, the outer automorphism group is described explicitly in terms of these clique-local pieces. The relative feature here is not boundary or subgroup preservation in the RAAG sense, but localization at clique subalgebras and their intersections.

For rigid geometries on leaf spaces of foliations, the relevant automorphism group is formed by diffeomorphisms of the desingularized basic manifold that preserve the rigid structure and commute with the $f$16-action. If $f$17 lies in the category $f$18 of leaf manifolds admitting a rigid geometry with zero structural Lie algebra, then $f$19 carries a unique finite-dimensional Lie group structure, with

$f$20

where $f$21 and $f$22 (Zhukova, 2017). This extends orbifold-type rigidity to leaf spaces that may be non-Hausdorff or fail the $f$23 axiom.

A different kind of relative rigidification appears in the theory of simplicial complexes. Any finite connected simplicial complex $f$24 can be rigidified to a new complex $f$25 with $f$26 for any prescribed subgroup $f$27, while preserving the homotopy type $f$28 (Costoya et al., 11 Sep 2025). The same construction realizes any action of a finite group on a finitely presentable group as the action of the self-homotopy equivalence group on a fundamental group. Although the terminology is somewhat different, this is again a rigid-relative principle: one starts with an ambient automorphism group and suppresses all automorphisms outside a chosen subgroup.

Taken together, these results show that rigidity is best viewed as a structural constraint on automorphism theory rather than a uniform property of a single category. In algebraic geometry it usually forces torus-dominated automorphism groups; in topology and representation theory it converts relative automorphism groups into finite extensions of mapping class groups; in fusion and linking systems it reduces rigid automorphisms to inner ones except for explicit $f$29-local exceptions; and in RAAG theory it makes the class of relative automorphism groups stable under restriction and decomposition. This suggests a broad mathematical principle: once the relevant peripheral data is fixed, rigidity tends to replace large, poorly controlled automorphism groups by explicit extensions, split sequences, or inductive hierarchies.