---
title: Boundary Tube Algebra in Theoretical Physics
url: https://www.emergentmind.com/topics/boundary-tube-algebra
type: topic
---

# Boundary Tube Algebra in Theoretical Physics

Searching arXiv for recent papers on boundary tube algebra and related tube algebra frameworks.
Boundary tube algebra denotes a family of boundary-localized annular or cut-local algebras that arise when generalized symmetry data, topological line data, or module-category data are restricted to a boundary, entanglement cut, interval endpoint, or gapped interface. In the cited literature, these algebras are constructed from boundary-compatible endpoint and junction operators, boundary lasso operators, or minimal tube diagrams, and they serve as the canonical receptacles for boundary sector information. Their center, primitive idempotents, or representation theory then organizes the measurable or topological sectors available at the boundary: in operational many-body settings this yields the finest admissible distinguishability data, while in SymTFT, string-net, and Turaev–Viro settings it classifies boundary-changing sectors, boundary excitations, and domain-wall excitations [2606.19678, 2409.02159, 2507.01515].

## 1. Definitions and scope

The term is used in several adjacent frameworks. In the operational setting of generalized symmetries, one begins with a generalized symmetry category \(\mathcal C\) acting on the entangling-cut boundary module \(M_A\), and the action defines a boundary tube algebra \(\mathrm{Tube}_{\mathcal C}(M_A)\). In SymTFT treatments of \(1+1\)d QFT, the relevant algebra is the boundary tube algebra \(\mathrm{Tube}(\mathcal B_1^\vee\vert \mathcal B_2)\), generated by boundary lasso operators acting on boundary-changing local operators. In Turaev–Viro and generalized multifusion string-net models, the boundary tube algebra is attached to a module category \(M\) describing a gapped boundary, and is written \(\mathrm{Tube}_{\rm bd}({}_C M)\) or \(\tilde{\mathbf{L}({_D M})}\) depending on conventions [2606.19678, 2409.02806, 2507.01515, 2403.04446].

| Setting | Notation | Role |
|---|---|---|
| Entanglement cut with generalized symmetry | \(\mathrm{Tube}_{\mathcal C}(M_A)\) | Cut-local endpoint/junction algebra |
| SymTFT interval boundary data | \(\mathrm{Tube}(\mathcal B_1^\vee\vert \mathcal B_2)\) | Algebra of boundary lasso operators |
| Gapped boundary of Turaev–Viro or string-net phase | \(\mathrm{Tube}_{\rm bd}({}_C M)\), \(\tilde{\mathbf{L}({_D M})}\) | Weak Hopf symmetry classifying boundary excitations |

These constructions are not merely changes of basis on a Hilbert space. They are defined by boundary compatibility conditions: locality at the cut, compatibility with the boundary module, or the existence of boundary junction data. A recurrent feature is that the algebra is semisimple in the finite cases under discussion, so its center and primitive idempotents provide canonical sector labels. Another recurrent feature is that the boundary algebra refines coarser observables or sector decompositions: scalar overlap data are refined by symmetry-resolved data, and symmetry-resolved data are further refined by tube-sector data; similarly, bulk tube algebra data are refined near a boundary by boundary-interface data [2606.19678, 2409.02159].

## 2. Operational tube sectors at entanglement cuts

In the operational tube-sector theory of quantum state distinguishability, the boundary tube algebra is the central kinematic object because it captures, at an entanglement cut, all generalized-symmetry endpoint data that remain physically measurable under the admissibility constraints. The reduced overlap operator is introduced as
\[
O_A=\sqrt{P_A O_A P_A},
\]
but it is not itself treated as an observable. The problem is instead to determine which sector readouts of the positive statistics \(\operatorname{Tr}(P\,O_A)\) are physically admissible when generalized symmetry lines may terminate on the cut. The answer is that if \(\mathcal C\) is the generalized symmetry category and \(M_A\) is the boundary module selected by the cut, then the action of \(\mathcal C\) on \(M_A\) defines the boundary tube algebra \(\mathrm{Tube}_{\mathcal C}(M_A)\), and the unique maximal commutative measurement algebra is its center:
\[
\mathcal A_{\mathrm{phys}} = Z\!\left(\mathrm{Tube}_{\mathcal C}(M_A)\right).
\]
The admissibility assumptions are complete positivity, entanglement-cut locality, boundary-module covariance, and sequential stability. Under these assumptions, every physically allowed sector readout is a coarse graining of the primitive central idempotents of the center, and every admissible measurement factors through \(\mathcal A_{\mathrm{phys}}\) followed by free classical postprocessing [2606.19678].

If \(e_a\) is a primitive central idempotent, its representation on the subsystem Hilbert space gives the projector
\[
P_a = R_A(e_a),
\]
and the normalized tube-sector probabilities are
\[
p_a=\frac{\operatorname{Tr}(P_a O_A)}{\operatorname{Tr}O_A},\qquad \sum_a p_a=1.
\]
These probabilities refine both scalar positive-overlap data and ordinary symmetry-resolved data. The hierarchy stated in the paper is
\[
\text{tube-sector resolution} \;\to\; \text{ordinary symmetry resolution} \;\to\; \text{scalar overlap}.
\]
For a forgetful map \(T:\mathcal T\to\mathcal Q\) from tube labels to coarser symmetry labels,
\[
p_q=\sum_{a:T(a)=q} p_a,\qquad p_{a|q}=p_a/p_q.
\]
The conditional distribution within a fiber \(T^{-1}(q)\) is precisely the information lost under coarse graining. This loss is operationally nontrivial because the restricted one-shot success probability for equal priors is controlled by total variation distance on the readout distributions, so tube-sector readout maximizes distinguishability among all admissible cut-local symmetry-compatible measurements. The paper’s doubled-Ising product–Kramers–Wannier example makes this concrete: the coarse \(G=\mathbb Z_2\times\mathbb Z_2\) resolution is blind to the split \(T^{-1}(0)=\{N_+,N_-\}\), while the tube readout detects changes in the conditional distribution on that fiber [2606.19678].

## 3. Boundary lasso algebras in SymTFT and interval QFT

In SymTFT treatments of finite, possibly non-invertible symmetry in \(1+1\)d QFT, the correct interval algebra is not the ordinary tube algebra of the \(S^1\) Hilbert space but a boundary tube algebra built from boundary lasso operators. The setting begins with a QFT \(Q\) with symmetry fusion category \(\mathcal C\), embedded into a Turaev–Viro theory \(\mathrm{TV}_{\mathcal C}\) with canonical regular boundary \(\mathcal B_{\rm reg}\) and physical boundary \(\widetilde Q\). Ordinary boundary conditions are decomposed as
\[
B \leftrightharpoons (\underline{B},\mathcal B,\widetilde B),
\]
where \(\mathcal B\) is a topological boundary of the SymTFT, \(\underline B\) is a topological line interface to \(\mathcal B_{\rm reg}\), and \(\widetilde B\) is a conformal interface to \(\widetilde Q\). For an interval with boundary conditions \(B_1,B_2\), the Hilbert space decomposes as
\[
\mathcal H_{B_1B_2}=\bigoplus_\rho \mathcal H^\rho_{B_1B_2}.
\]
The generators of the boundary tube algebra are boundary lasso operators
\[
\mathsf{H}^{C_1 C_2, y_1 y_2}_{B_1B_2,c},
\]
obtained by wrapping a topological line \(c\in\mathcal C\) around a boundary-changing operator and shrinking it onto the boundary junction. These operators generate
\[
\mathrm{Tube}(\mathcal B_1^\vee\vert \mathcal B_2),
\]
and their multiplication has the schematic form
\[
\mathsf H_{X'}\times \mathsf H_X=\sum_Y t_{X'X}^Y\,\mathsf H_Y,
\]
with structure constants determined by boundary \(\widetilde F\)-symbols. When \(\mathcal B_1=\mathcal B_2\), the boundary tube algebra becomes a \(C^\ast\)-weak Hopf algebra [2409.02806, 2409.02159].

The representation theory is the interval analogue of the usual bulk tube-algebra story. The irreducible representations of \(\mathrm{Tube}(\mathcal B_1^\vee\vert \mathcal B_2)\) are in one-to-one correspondence with simple topological line interfaces between \(\mathcal B_1\) and \(\mathcal B_2\), equivalently with simple objects of \(\mathrm{Fun}_{\mathcal C}(\mathcal B_1,\mathcal B_2)\). Projectors onto the \(\rho\)-sectors are written as linear combinations of lasso operators with coefficients determined by generalized half-linking numbers. This structure is then used to refine annulus partition functions and derive the noninvertible symmetry-resolved Affleck–Ludwig–Cardy formula and the associated symmetry-resolved entanglement entropy of a single interval [2409.02806, 2409.02159].

The critical double Ising model provides a concrete example. For a strongly symmetric interval boundary condition preserving the product of two Kramers–Wannier symmetries, the boundary lasso operators realize the Kac–Paljutkin Hopf algebra \(H_8\). The invertible lines \(g\in\mathbb Z_2\times\mathbb Z_2\) give operators \(\mathsf H_g\), while the noninvertible line \(\mathcal N\) gives four operators \(\mathsf H_{\mathcal N}^{ss'}\). Their multiplication rules reproduce the \(H_8\) algebra, and the corresponding sector-resolved entropies distinguish the noninvertible sector from the invertible ones by their universal constant terms [2409.02806].

## 4. Weak Hopf boundary tube algebras in string-net and Turaev–Viro theories

In Levin–Wen, Turaev–Viro, and generalized multifusion string-net settings, boundary tube algebra is the algebraic object associated with a gapped boundary specified by a module category. If the bulk phase is encoded by a fusion category \(C\) and the boundary by a left module category \({}_C M\), the notation
\[
\mathrm{Tube}_{\rm bd}({}_C M)
\]
is used. Conceptually, this is the algebra of operators acting on a minimal tubular neighborhood around a boundary excitation. The construction employs module-category local moves—loop move, parallel move, and \(F\)-move—to reduce boundary-tube diagrams to a basis of minimal tubes. In the generalized multifusion string-net model, analogous left and right boundary tube algebras are denoted \(\tilde{\mathbf{L}({_D M})}\) and \(\tilde{\mathbf{R}(M_D)}\), and their modules classify boundary excitations [2507.01515, 2403.04446].

A major structural point is that these are not only associative algebras but weak Hopf algebras. The multiplication is defined by stacking tubes and reducing the result by \(F\)-moves and loop or parallel moves; the unit is the identity tube; the counit collapses a tube to a scalar; the coproduct inserts intermediate labels and sums over channels; the antipode turns the tube around; and under the condition
\[
d_m^L=d_m^R=d_m,
\]
the involution defined by reflection gives a \(C^\ast\)-weak Hopf algebra. The domain-wall tube algebra \(\mathbf{Tube}({}_C M_D)\) is the two-sided analogue for a bimodule category \({}_C M_D\), and the paper states that it is a \(C^\ast\) weak Hopf algebra with these structure maps. Boundary and domain-wall excitations are then classified by representations of the corresponding tube algebras [2507.01515].

The relation between boundary and domain-wall objects is especially explicit. One paper establishes a skew-pairing
\[
p:\mathrm{Tube}_{\rm bd}({M}_D)\times \mathrm{Tube}_{\rm bd}({}_C M)\to \mathbb C
\]
and proves the weak-Hopf isomorphism
\[
\mathbf{Tube}({}_C M_D)\simeq \mathrm{Tube}_{\rm bd}({M}_D)\Join_p \mathrm{Tube}_{\rm bd}({}_C M).
\]
Another paper shows that in the generalized multifusion string-net model the bulk tube algebra is a crossed product of left and right boundary tube algebras, and that the domain-wall tube algebra is similarly obtained from the corresponding boundary algebras. In both formulations, the boundary tube algebra is the basic one-sided object from which the doubled or two-sided wall algebra is assembled. The folding trick and Morita-theoretic arguments explain why a domain wall may be reinterpreted as a boundary for a folded theory [2507.01515, 2403.04446].

## 5. Categorical tube algebras, Drinfeld centers, and localized corners

The boundary tube algebra is best understood against the background of the ordinary tube algebra of a rigid tensor category. For a rigid \(C^\ast\)-tensor category \(\mathcal C\) with simple representatives \(U_i\), the tube algebra is
\[
\mathrm{Tub}(\mathcal C)=\bigoplus_{i,j,s\in \mathrm{Irr}(\mathcal C)} \mathrm{Hom}(U_s\otimes U_j,\; U_i\otimes U_s),
\]
with multiplication given by gluing annuli and involution determined by duality. Its representation category is equivalent to the Drinfeld center \(Z(\mathrm{ind}\text{-}\mathcal C)\). In the case of a spherical semisimple multitensor category, the category of tube algebra representations carries a braided monoidal structure with twist and is braided tensor equivalent to the Drinfeld center of the ind-category [1511.06332, 2511.08482].

Within this general theory, several papers isolate sector-localized pieces that function as natural precursors or analogues of boundary tube algebras. For a \(\Gamma\)-graded rigid \(C^\ast\)-tensor category, the tube algebra decomposes as a Fell bundle over the adjoint-action groupoid \(\Gamma\ltimes_{\mathrm{Ad}}\Gamma\). Each conjugacy-class component is Morita equivalent to a centralizer corner \(A_{a,a}\), and twisting the associator by a normalized \(\mathbb T\)-valued \(3\)-cocycle induces a \(2\)-cocycle twist on these corners. The paper does not formally define a boundary tube algebra, but it identifies the corners \(A_{a,a}\) as the natural localized algebras attached to a sector or stabilizer [1801.01735].

A closely related decomposition appears for tube algebras of diagonal and Bisch–Haagerup subfactors with scalar \(3\)-cocycle obstruction. There the representation theory decomposes over conjugacy classes, and each block is a matrix algebra tensored with a twisted group algebra of the corresponding centralizer. In formula form,
\[
\mathrm{Rep}(\text{tube algebra}) \simeq \prod_{[g]\subseteq G} \mathrm{Rep}\big(G_g,\varphi_g\big),
\]
where \(G_g\) is the centralizer of \(g\) and
\[
\varphi_g(x,y)=\overline{\omega}(g,x,y)\,\omega(x,g,y)\,\overline{\omega}(x,y,g).
\]
This blockwise centralizer structure is not itself called a boundary tube algebra in the paper, but it provides the same kind of sector decomposition that later boundary-localized constructions exploit [1701.00097].

## 6. Related notions, invariance results, and conceptual limits

The term boundary tube algebra is not completely standardized, and nearby literature contains both genuine analogues and explicit non-examples. In the theory of \(GL_m\)-dimers attached to triangulations of a convex \(n\)-gon, the relevant object is the boundary algebra
\[
\mathcal B_Q:=e_b\Lambda_Q e_b,
\]
where \(e_b\) is the sum of boundary idempotents and \(\Lambda_Q\) is the dimer algebra. The paper proves that for any two triangulations \(T_1,T_2\) of the same \(n\)-gon, the corresponding boundary algebras are isomorphic. It consistently uses the term boundary algebra rather than boundary tube algebra, but it explicitly states that the structure may be interpreted as the boundary algebra or tube algebra attached to the triangulation class, because it is the triangulation-independent boundary endomorphism algebra [1804.07243].

By contrast, some boundary-centered frameworks are only structurally analogous. In “Algebraic Phase Theory V,” the main boundary objects are the structural boundary, the canonical exact sequence
\[
0 \longrightarrow \mathcal R \longrightarrow \mathcal P \longrightarrow \mathcal B \longrightarrow 0,
\]
the boundary quotient \(\mathcal B=\mathcal P/\mathcal R\), and the rigidity island \(\mathcal I_{\mathcal P}\). The paper explicitly states that it does not develop a tube algebra in the usual algebraic sense: it does not define multiplication rules, annular composition laws, or a tube-category endomorphism algebra. Its boundary quotient is a universal obstruction object, and the framework is described as boundary calculus, obstruction calculus, and rigidity-island formalism rather than tube algebra [2601.18257].

A plausible implication of these contrasts is that “boundary tube algebra” should be reserved for boundary-localized algebras that retain genuine tube-algebra features: annular or cut-local composition, canonical sector idempotents or irreducible modules, and a direct link to boundary measurements, boundary-changing operators, or boundary excitations. On that narrower reading, the operational entanglement-cut theory, the SymTFT boundary-lasso theory, and the weak Hopf boundary theories of string-net and Turaev–Viro models provide the clearest current realizations [2606.19678, 2409.02159, 2507.01515].

Source: https://www.emergentmind.com/topics/boundary-tube-algebra