[Paper Review] On a subfactor generalization of Wall's conjecture

Robert M. Guralnick, Feng Xu|arXiv (Cornell University)|Jun 30, 2010

Advanced Operator Algebra Research15 references4 citations

TL;DR

This paper proposes a subfactor generalization of Wall's conjecture on maximal subgroups in finite groups, establishing it for solvable groups and Kac algebras of Izumi-Kosaki type. It proves that the number of maximal intermediate subfactors is bounded by the dimension of the second relative commutant, using double coset counting and group-theoretic techniques, and formulates a tensor product conjecture for subfactors with applications to finite group realizations.

ABSTRACT

In this paper we discuss a conjecture on intermediate subfactors which is a generalization of Wall's conjecture from the theory of finite groups. We explore special cases of this conjecture and present supporting evidence. In particular we prove special cases of this conjecture related to some finite dimensional Kac Algebras of Izumi-Kosaki type which include relative version of Wall's conjecture for solvable groups.

Motivation & Objective

• To generalize Wall’s conjecture on maximal subgroups in finite groups to the setting of subfactors in von Neumann algebras.
• To investigate whether the number of maximal intermediate subfactors in a finite-index irreducible subfactor is bounded by the dimension of the second relative commutant.
• To extend the conjecture to tensor products of subfactors and verify it in special cases using group-theoretic and cohomological tools.
• To provide evidence for the conjecture through explicit constructions in Kac algebras and solvable group settings.

Proposed method

• Use of the commutant map to relate intermediate subfactors in $N \subset M$ to those in $M \subset M_1$, preserving lattice structure and duality.
• Application of double coset counting arguments to bound the number of maximal subgroups containing a given subgroup $H$ in finite groups.
• Leveraging the cross-product construction to realize group-theoretic interval lattices as subfactor lattices.
• Employing the Aschbacher-O’Nan-Scott theorem to analyze primitive quotients of group extensions and classify maximal subgroups in direct products.
• Using the classification of finite simple groups to bound $|\mathrm{Aut}(S)|$ for non-abelian simple groups $S$, ensuring $|\mathrm{Aut}(S)| < (|S|-1)^2$.
• Formulating a tensor product conjecture where the number of non-trivial minimal intermediate subfactors in $M_1 \otimes M_2$ is bounded by $(n_1 - 1)(n_2 - 1)$, with $n_i = \dim(N_i' \cap M_{i,1})$.

Experimental results

Research questions

• RQ1 Can Wall’s conjecture on maximal subgroups in finite groups be generalized to the setting of subfactors in von Neumann algebras?
• RQ2 Is the number of maximal intermediate subfactors in a finite-index irreducible subfactor $N \subset M$ bounded by the dimension of the second relative commutant $N' \cap M_1$?
• RQ3 What is the structure of intermediate subfactors in tensor products of subfactors, and can a uniform bound be established for non-trivial components?
• RQ4 How do Kac algebras of Izumi-Kosaki type realize the generalized Wall’s conjecture, particularly in the solvable case?
• RQ5 Does the tensor product conjecture hold for subfactors arising from finite groups, especially when $X$ and $Y$ are elementary abelian 2-groups?

Key findings

• The relative version of Wall’s conjecture holds for finite groups when $H$ is a subgroup of $G$, with the number of maximal subgroups strictly containing $H$ bounded by the number of double cosets of $H$ in $G$.
• For solvable groups, the maximal and minimal versions of the generalized Wall’s conjecture are verified via counting arguments and coideal mappings in Kac algebras.
• The number of maximal subgroups in $G = X \times Y$ that contain neither $X$ nor $Y$ is at most $(x-1)(y-1)$, with equality if and only if $X$ and $Y$ are elementary abelian 2-groups.
• The tensor product conjecture is proposed: the number of minimal intermediate subfactors in $M_1 \otimes M_2$ not of the form $N_1 \otimes P$ or $P \otimes N_2$ is at most $(n_1 - 1)(n_2 - 1)$, where $n_i = \dim(N_i' \cap M_{i,1})$.
• The proof of Lemma 4.1 relies on the classification of finite simple groups to bound $|\mathrm{Aut}(S)| < (|S|-1)^2$, which ensures the inequality holds in the direct product case.
• The conjecture is verified in special cases, including when $G = X \times Y$ with $X,Y$ elementary abelian 2-groups, and when $G$ is solvable, showing consistency with the generalized Wall’s conjecture.