The zipper condition for 4 4 -tensors in two-dimensional topological order and the higher relative commutants of a subfactor arising from a commuting square
Yasuyuki Kawahigashi · 2025
Read The zipper condition for 4 4 -tensors in two-dimensional topological order and the higher relative commutants of a subfactor arising from a commuting square by Yasuyuki Kawahigashi (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-2511-10080.txt.
Introduction
Yasuyuki Kawahigashi, writing alone, connects two languages for describing two-dimensional topological order: the tensor networks used by condensed matter researchers and the subfactor theory of operator algebras. Physicists' four-tensors in these networks were expected to match bi-unitary connections in Jones' theory. Kawahigashi makes that identification precise, including the exact normalization constants.
He then proves that two-tensors satisfying the zipper condition from physics are exactly the flat fields of strings in that theory, which correspond to elements of the higher relative commutants. Notably, the proof needs neither flatness nor finite depth of the connection, and it generalizes the tensors so all four index sets may differ. The bridge lets results travel between physics and algebra.
Catalog record
| CREATOR | Yasuyuki Kawahigashi |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 25 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2511-10080
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
The zipper condition for 4 4 -tensors in two-dimensional topological order and the higher relative commutants of a subfactor arising from a commuting square Yasuyuki Kawahigashi Abstract They recently study two-dimensional topological order in condensed matter physics in terms of tensor networks involving certain 3 3 - and 4 4 -tensors. Their 3 3 -tensors satisfying the “zipper condition” play an important role there. We identify their 4 4 -tensors with bi-unitary connections in Jones’ subfactor theory in operator algebras with precise normalization constants. Then we prove that their tensors satisfying the zipper condition are the same as flat fields of strings in subfactor theory which correspond to elements in the higher relative commutants of the subfactor arising from the bi-unitary connection. This is what we expect, since the zipper condition is a kind of pentagon relations, but we clarify what conditions are exactly needed for this — we do not need the flatness or the finite depth condition for the bi-unitary connection. We actually generalize their 4 4 -tensors so that the four index sets of the 4 4 -tensors can be all different and work on a “half-version” of the zipper condition. Dedicated to the memory of Hikosaburo Komatsu 1 Introduction Fusion categories [ 5 ] (with or without braiding) have emerged as new types of symmetries in mathematics and physics. Both quantum field theory and condensed matter physics have seen such symmetries recently and they are often called non-invertible symmetries in the physics literature, and also called quantum symmetries in various mathematics literature. Particularly, many researchers in two-dimensional topological order in condensed matter physics are interested in such studies using tensor networks recently as in [ 2 ] , [ 24 ] , [ 29 ] . It has been well-known that subfactor theory of Jones [ 13 ] , [ 14 ] in operator algebras gives appropriate tools to study structures of fusion categories, and it is indeed this theory which led to the discovery of the Jones polynomial , the first mathematical realization of quantum symmetry. This approach is closely related to operator algebraic studies of quantum field theory [ 21 ] , [ 22 ] , [ 7 ] , since early days of subfactor theory through Doplicher-Haag-Roberts theory in algebraic quantum field theory . In a usual operator algebraic study of fusion categories, we realize an object of a fusion category as a bimodule over (type II 1 ) factors or an endomorphism of a (type III) factor [ 6 ] , [ 7 ] . Another approach [ 1 ] based on bi-unitary connections [ 25 ] , [ 31 ] , [ 15 ] is less common, but contains the same information as these two methods and has an advantage that everything is finite dimensional . This finite dimensionality enabled us to construct the Haagerup subfactor [ 1 ] , which is still one of the most mysterious quantum symmetries today. Recall that a bi-unitary connection gives a characterization [ 31 ] , [ 6 ] of a non-degenerate commuting square, which was initially studied in [ 27 ] in a different context and has complete information to recover an amenable subfactor of type II 1 [ 28 ] . It has been pointed out in [ 16 ] , [ 18 ] that the 4 -tensors in [ 2 ] are mathematically the same as bi-unitary connections up to slight change of normalization, and identification of some natural Hilbert spaces in condensed matter physics and subfactor theory has been given in [ 17 ] . A characterization of such 4 -tensors as certain generalized quantum 6j -symbols has been also given in [ 19 ] . This shows that anyons [ 20 ] are studied with such 4 -tensor networks [ 9 ] . Our correspondence among various mathematical approaches to study fusion categories are summarized in Tables 1 and 2 . Table 1: Correspondence among endomorphisms, bimodules and connections endomorphism bimodules connections identity identity bimodule trivial connection direct sum direct sum direct sum composition relative tensor product composition conjugate endomorphism dual bimodule dual connection dimension (\text{Jones\ index})^{1/2} Perron-Frobenius eigenvalue intertwiner intertwiner flat field of strings Table 2: Correspondence between connections, commuting squares and 4 -tensors connections commuting square 4 -tensor trivial connection commuting square trivial 4 -tensor direct sum direct sum direct sum composition composition concatenation dual connection basic construction complex conjugate tensor Perron-Frobenius eigenvalue (\text{Pimsner-Popa\ index})^{1/2} Perron-Frobenius eigenvalue flat fields of strings relative commutant tensors with the zipper condition The aim is this paper is threefold and to complete the above Tables as follows. (1) Give precise normalization constants in various formulas. (2) Characterize the morphism property in terms of the “if and only if” form in comparison to the open string bimodule framework and the zipper condition [ 2 , (2)] . (3) Give the most general setting of assumptions under which our arguments work. The first one is only a technical issue, but important for actual computations. Examples of concrete computations in various papers arise from 3 -cocycles on finite groups, where all the normalizing constants are 1 and this issue can be ignored, but we need them in a more general setting. The second is important from a theoretical viewpoint. The bimodule approach involves infinite dimensional operator algebras and Hilbert spaces, so it is not clear whether this approach gives the same morphisms as in the tensor network framework. For the third aim, we may have four different index sets for our 4 -tensors, as long as we have bi-unitarity as in Fig. 26 and 27 , and we do not need the finite depth condition or the flatness condition [ 6 ] for bi-unitary connections. The lack of the former condition means that the initial data can produce countably many irreducible objects in our tensor category, and the lack of the latter condition means that our bi-unitary connection, a kind of quantum 6j -symbols, does not have to be in a canonical form. 2 Bi-unitary connections We prepare notations and conventions on bi-unitary connections as in [ 1 ] , [ 6 , Chapter 11] , [ 15 ] , [ 25 ] , [ 31 ] . We give complete definitions since our setting is slightly more general than the one in [ 16 ] , [ 17 ] . We have four finite bipartite oriented graphs {\mathcal{G}}_{0},{\mathcal{G}}_{1},{\mathcal{G}}_{2},{\mathcal{G}}_{3} . We assume that {\mathcal{G}}_{0} and {\mathcal{G}}_{2} are connected. Each graph can have multiple edges between one pair of vertices and also cycles, but cannot have a loop, an edge from one vertex to the same one, since it is bipartite. We write E({\mathcal{G}}) for the edge set of a graph {\mathcal{G}} . We assume that the sets of the source vertices of E({\mathcal{G}}_{0}) , E({\mathcal{G}}_{1}) , E({\mathcal{G}}_{2}) and E({\mathcal{G}}_{3}) are V_{0} , V_{0} , V_{1} and V_{3} , respectively. We further assume that the sets of the range vertices of E({\mathcal{G}}_{0}) , E({\mathcal{G}}_{1}) , E({\mathcal{G}}_{2}) and E({\mathcal{G}}_{3}) are V_{3} , V_{1} , V_{2} and V_{2} , respectively. We draw a diagram as in Fig. 1 to depict this situation. We assume that the numbers of the edges of all the four graphs are larger