Inclusions of Operator Algebras from Tensor Categories: beyond irreducibility
Lucas Hataishi, Roberto Hernández Palomares · 2025
Read Inclusions of Operator Algebras from Tensor Categories: beyond irreducibility by Lucas Hataishi, Roberto Hernández Palomares (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2504-05247.txt.
Introduction
Lucas Hataishi and Roberto Hernandez Palomares build inclusions of algebras from tensor categories, in a 2025 paper on math. Their build uses a coend-type recipe inside unitary tensor categories to make faithful inclusions of C-star algebras. Older theory leaned on irreducibility, a trait that many natural cases lack, so the old frame left those cases out.
The work sets out more than one notion of discreteness for such inclusions when that trait is gone, all in one shared frame. The frame also covers actions of compact quantum groups, so a wide set of cases sits under one roof. Cases come from semi-circular systems and from factorization homology. When the trait does hold, the authors pin down when two discreteness tests agree, which tidies a corner of operator algebra theory.
Catalog record
| CREATOR | Lucas Hataishi, Roberto Hernández Palomares |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 128 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2504-05247
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Inclusions of Operator Algebras from Tensor Categories: beyond irreducibility Lucas Hataishi, Roberto Hernández Palomares Abstract We derive faithful inclusions of C*-algebras from a coend-type construction in unitary tensor categories. This gives rise to different potential notions of discreteness for an inclusion in the non-irreducible case, and provides a unified framework that encloses the theory of compact quantum group actions. We also provide examples coming from semi-circular systems and from factorization homology. In the irreducible case, we establish conditions under which the C*-discrete and W*-discrete conditions are equivalent. 1 Introduction Inclusions of operator algebras provide a flexible framework to study classical and quantum symmetries of noncommutative spaces. Jones’ landmark discovery of the Index Rigidity Theorem for subfactors [ Jon83 ] launched the construction and classification of numerous new examples of subfactors, and fueled interactions with other fields such as low-dimensional topology, statistical mechanics, and quantum field theory. (See [ JS97 ] for a survey). Subsequently, discrete subfactors were introduced in [ ILP98 ] as a broad class of inclusions N\subset M sharing properties of crossed products by outer actions of discrete groups on factors. They generally do not have finite index, but can be characterized either by a generating property in terms of the quasi-normalizer of N in M , or by a decomposition of the space of bounded N{\hbox{-}}N -bimodular operators on L^{2}(M) as a product of type I factors. Associated to a subfactor N\subset M is its standard invariant , {\mathcal{C}}_{N\subset M} , that can be axiomatized using Jones’ planar algebras [ Jon22 ] or Popa’s \lambda -lattices [ Pop95 ] . The standard invariant classifies amenable subfactors of the hyperfinite \rm II_{1} -factor \mathcal{R} [ Pop94 ] . The formulation of the standard invariant that is relevant to us follows the philosophy advocated in [ JP17 , Müg03a , Müg03b ] , and goes roughly as follows. Let N be a a \II_{1} -factor N , and denote by {\sf Bim}(N) the category of N{\hbox{-}}N -Hilbert bimodules. Suppose we are given a subfactor N\subset M . Consider the Hilbert space completion L^{2}(M) with respect to the trace. Looking at it as a N{\hbox{-}}N -bimodule, it decomposes into a direct sum of irreducible ones. Let {\mathcal{C}}_{N\subset M} be the subcategory of {\sf Bim}(N) generated by those irreducible bimodules, their direct sums and their fusion products. By construction, {\mathcal{C}}_{N\subset M} is a C ∗ -tensor subcategory of {\sf Bim}(N) . The factor M induces a functor {\mathbb{M}}:({\mathcal{C}}_{N\subset M})^{\op}\to{\sf Vec} , defined on objects by {\mathbb{M}}(K)=\Hom_{N-M}(K\underset{N}{\boxtimes}L^{2}(M),L^{2}(M)) . Then the standard invariant of N\subset M is the pair ({\mathcal{C}}_{N\subset M},{\mathbb{M}}) . The main result in [ JP19 ] is that the standard invariants ({\mathcal{C}}_{N\subset M},{\mathbb{M}}) classify extremal irreducible discrete extensions N\subset M . We say that \ {\mathcal{C}}_{N\subset M} acts on N by means of the inclusion functor {\mathcal{C}}_{N\subset M}\hookrightarrow{\sf Bim}(N) . The procedure by which M is reconstructed from the standard invariant is called the realization or crossed product of N by {\mathbb{M}} , and it is denoted by N\rtimes{\mathbb{M}} . We say that {\mathbb{M}} is the C ∗ -algebra object , in the algebraic ind-completion {\sf Vec}({\mathcal{C}}_{N\subset M}):=\{\text{linear functors}\ F:{\mathcal{C}}_{N\subset M}^{\op}\to{\sf Vec}\} of {\mathcal{C}}_{N\subset M} , associated to the subfactor N\subset M . In [ HPN23 ] , the second-named author and Nelson introduced the abstract properties of discreteness and projective-quasi-regularity for unital irreducible inclusions of C*-algebras A\overset{E}{\subset}B . Akin to discrete subfactors, irreducible C*-discrete inclusions are characterized by a standard invariant comprising an action of a UTC and a connected C*-algebra object in its algebraic ind-completion. The class of C*-discrete inclusions encompasses all finite Watatani-index, certain semicircular systems, and cores of Cuntz algebras, as well as crossed products by actions of discrete (quantum) groups. A main objective of this manuscript is to extend the standard invariant to non-irreducible inclusions of C*-algebras. A prominent class of examples comes from the Morita theory of compact quantum group actions, the relevant inclusions beings B^{\mathbb{G}}\subset B , where B is a C*-algebra acted by a compact quantum group and B^{\mathbb{G}} is the fixed point C*-subalgebra. Such inclusions have a well understood categorical description based on Tannaka-Krein duality [ DCY13 , DCY15 , Nes14 ] . Such an approach to quantum group actions has been very useful in classification problems, e.g. [ NY17 ] . In this context, irreducibility corresponds to ergodicity of the action of the compact quantum group. Cores of Cuntz algebra fall into this class of examples. Another source of motivation for this work comes from factorization homology with values in C*-categories (cf [ Hat23 ] ). Starting with a unitarily braided unitary tensor category {\mathcal{C}} , factorization homology gives a functorial construction associating a C*-tensor category F_{\mathcal{C}}(\Sigma) to every compact oriented surface \Sigma . When \Sigma has non-empty boundary, F_{\mathcal{C}}(\Sigma) is described by a C ∗ -algebra object that is generally not connected. We shall now describe the contents of this article. Section 2 is a preliminary section where we gather some background material, fix notations and conventions. In Section 3 we introduce the coend realization construction that will serve as main tool in this paper. Given a UTC {\mathcal{C}} , C*-algebra objects {\mathbb{A}} in {\sf Vec}({\mathcal{C}}^{\op}) and {\mathbb{D}} in {\sf Vec}({\mathcal{C}}) , it follows from [ JP19 , §4.4] , the existence of a C*-algebra C^{*}_{u}|{\mathbb{A}}\odot{\mathbb{D}}| . We show that there is also a minimal completion, which we denote by {\mathbb{A}}\bowtie{\mathbb{D}} , and call it the coend realization of the pair ({\mathbb{A}},{\mathbb{D}}) . Let \mathbbm{1} be a tensor unit in {\mathcal{C}} . We then prove the following. Theorem A (Theorem 3.15 ) . There is a canonical faithful conditional expectation {\mathbb{E}}:{\mathbb{A}}\bowtie{\mathbb{D}}\to{\mathbb{A}}(1)\otimes{\mathbb{D}}(\mathbbm{1}) . The proof of the theorem relies on the existence of canonical expectations of finite Pimsner-Popa indices {\mathbb{E}}^{{\mathbb{A}}}_{X}:{\mathbb{A}}(\overline{X}X)\to{\mathbb{A}}(1) and {\mathbb{E}}^{\mathbb{D}}_{X}:{\mathbb{D}}(\overline{X}X)\to{\mathbb{D}}(\mathbbm{1}) , which allow for a local comparison between operator norms and Hilbert C*-module norms. We then consider the above construction to the case when {\mathbb{A}} is the action of a UTC on a unital C*-algebra A with trivial center. This means that {\mathbb{A}} is a unitary tensor functor {\mathbb{A}}:{\mathcal{C}}\to{\sf Bim_{fgp}}(A) from {\mathcal{C}} to the UTC of finitely generated projective bimodules of A . In this context, we write A\rtimes{\mathbb{D}} for {\mathbb{A}}\bowtie{\mathbb{D}} , and call it the crossed product of A by {\mathbb{D}} under the action of {\mathcal{C}} . Then, by applying Theorem A, since {\mathbb{A}}(1)=A , we obtain the inclusions A\subset A\otimes{\mathbb{D}}(\mathbbm{1})\overset{{\mathbb{E}}}{\subset}A\rtimes{\mathbb{D}}, where the last inclusion is faithful. It is then easy to see, assuming {\mathcal{Z}}(A)\simeq{\mathbb{C}} , that the possible descents of {\mathbb{E}}:A\rtimes{\mathbb{D}}\to A\otimes{\mathbb{D}}(\mathbbm{1}) to a conditional expectation E:A\rtimes{\mathbb{D}}\to A are parametrized by states on the C ∗ -algebra {\mathbb{D}}(\mathbbm{1}) . As mentioned earlier, in the case of a \rm II_{1} -factor N , a discrete extension N\subset M can be characterized in two ways; either in terms of the quasi-normalizer or in terms of \End_{N-N}(L^{2}(M)) being a discrete von Neumann algebras. For extensions of C ∗ -algebras, it is yet not known if the two, accordingly adapted to the C ∗ -context, are equivalent. Taking this into account, in Section 4, we make the distinction between flavors of discreteness for a faithful inclusion A\overset{E}{\subset}D . Using the faithful conditional expectation E , we can complete D to an A{\hbox{-}}A -correspondence {\mathcal{E}} . Discreteness and their variations are