{"id":"7876b2cd-988a-4fbb-bf0c-4c1474d73a40","arxiv_id":"2506.19969","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For topological boundaries in Levin-Wen and Walker-Wang models, the boundary algebra's DHR bimodule category recovers the boundary topological order, and for Walker-Wang it is the enriched center Z_B(X).","lead":"The authors extend a rigorous operator-algebra framework for topological order to systems with topological boundaries, and show that boundary excitations are captured by the boundary algebra's DHR bimodule category. The results put topological holography on a precise mathematical footing for Levin-Wen and Walker-Wang models, connecting quantum spin systems to subfactor theory and tensor categories.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The folding-trick step in §4.4 is unproved and load-bearing: the equivalence DHR^∂(X) ≅ DHR^∂(M) is asserted, and Theorems D and E depend on it.","rationale":"I read the full manuscript. The Levin-Wen analysis in §3 and Appendix A is largely independent of the folding trick and is built on established subfactor and Q-system realization machinery; that part is credible. The braided categorical net analysis in §4.1 proves Haag duality and identifies the W*-category of superselection sectors in Theorem 4.8, but explicitly defers fusion and braiding, as the paper's footnote 1 acknowledges. The central new claim for Walker-Wang boundaries, Theorem D, depends on the folding-trick step in §4.4, which is asserted in a single sentence and not proved. The same step is used to transport boundary algebraic Haag duality and to justify the braiding construction in Theorem 4.21. This is a genuine, load-bearing gap: the categorical equivalence DHR^∂(X) ≅ Z_A(X) is the core of the paper's holography claim, and without a proof of the folding bijection the theorem is not established. I do not see a reason to reject the framework outright; the gap is localized and plausibly fillable, so the reader's conditional verdict is appropriate and I do not recommend changing it. My proposed test either directly verifies the folding equivalence in a finite example or forces an explicit construction of the functors, which would settle whether the asserted identification holds.","tokens_in":39944,"tokens_out":8377,"duration_ms":83618,"concrete_test":"Construct the folded 1D net M from the 2D braided-enriched net X explicitly for a finite nontrivial example, e.g., A = Vec(Z/2) with the symmetric braiding and a concrete A-enriched fusion category X, and compute DHR^∂(X) and DHR^∂(M) directly from Definition 2.20 using the local triviality characterization of Lemma 2.21. If the two categories are inequivalent, the folding observation in §4.4 is false. Alternatively, prove the observation by writing down the two functors between DHR^∂(X) and DHR^∂(M) induced by the folding bijection, verifying that they are well-defined on localizable bimodules and inverse up to natural isomorphism; failure to construct either functor would pinpoint the missing argument.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim of Theorem D (and its corollary Theorem E) rests on the folding-trick identification in §4.4. The paper states, in one sentence, that 'any boundary DHR bimodule in DHR^∂(X) can be viewed as a boundary DHR bimodule in DHR^∂(M)', but no functor between these categories is constructed and no proof is given that localizability (Definition 2.20 and its equivalent form Lemma 2.21) is preserved under the bijection Z×N→N. The subsequent chain DHR^∂(M) ≅ End(M_C) (Theorem 3.14) and End(M_C) ≅ Z_A(X) (via [HBJP23, Ex. II.9]) therefore identifies the DHR category of the folded 1D net, not necessarily of the original 2D net. The braided equivalence in Theorem 4.21 inherits the same gap: boundary algebraic Haag duality is transferred from the 1D net (3.13) to the 2D net 'since' the folding trick applies, and the braidedness of the functor Y is only sketched in the A = Hilb case, with the general case dismissed by saying the A-strands 'play no role'. If the folding bijection adds or removes sectors, or changes localizability, then DHR^∂(X) is not Z_A(X) and Theorems D and E do not follow as stated. This is a localized gap in proof rather than a known counterexample; the conditional verdict is appropriate.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper extends the authors' earlier local topological order (LTO) framework to bulk-boundary systems. It introduces boundary LTO axioms for nets on lattices with a boundary, defines a category of boundary DHR bimodules, and applies the framework to Levin-Wen and Walker-Wang models. The main results are: Theorem A, identifying the boundary DHR bimodules of a Levin-Wen bulk-boundary system with End(M_C); Theorem D, identifying the boundary DHR bimodules of the Walker-Wang bulk-boundary system for B -> Z(X) with the enriched center Z_B(X); and Theorem E, identifying the bulk Walker-Wang DHR category with the M\\\"uger center Z_2(B). Along the way the paper introduces braided categorical nets and braided-enriched fusion categorical nets, studies their superselection sectors, and shows (Theorems B and C) that the cone von Neumann algebras in the canonical boundary state are type I with finite-dimensional centers and that the superselection sector W*-category is Hilb(B).","tokens_in":40311,"tokens_out":5289,"duration_ms":57741,"significance":"If the main theorems are correct, the paper gives a substantial operator-algebraic formalization of topological holography for gapped boundaries, recovering expected categorical data from microscopic spin models. The Levin-Wen analysis is firmly grounded in established subfactor and Q-system techniques, and Appendix A provides a detailed and useful construction of AF actions of unitary multifusion categories; the DHR bimodule construction is explicit and parameter-free. The braided categorical nets are also a new and interesting class of nets with nontrivial superselection theory, and the type-I cone algebra result contrasts nicely with the type II/III behavior found in earlier Levin-Wen examples. However, the central Walker-Wang results depend on a folding-trick identification in Section 4.4 that is asserted rather than proved, and the braiding verification in Theorem 4.21 is only carried out in detail for A = Hilb. These gaps are localized but load-bearing; once they are filled, the significance is high.","major_comments":[{"comment":"The statement that 'any boundary DHR bimodule in DHR^∂(X) can be viewed as a boundary DHR bimodule in DHR^∂(M)' is load-bearing for Proposition 4.19 and for Theorems D and E, but no functor DHR^∂(X) -> DHR^∂(M) is constructed, and no proof is given that localizability in the sense of Definition 2.20 and Lemma 2.21 is preserved by the 'careful bijection Z×N→N'. Without such a proof, the chain DHR^∂(X) ≅ DHR^∂(M) ≅ End(M_C) ≅ Z_A(X) identifies the DHR category of the folded 1D net, not necessarily of the original 2D net. This is the central claim of the paper, so the missing construction and proof must be supplied.","section":"§4.4 (paragraph after Definition 4.16)"},{"comment":"The paper asserts that because the 1D fusion module categorical net satisfies boundary algebraic Haag duality by (3.13), the 2D braided-enriched fusion categorical net also satisfies boundary algebraic Haag duality. This transfer is not automatic and is needed for Lemma 2.23 and for the braiding computation in Theorem 4.21. The authors should either prove the 2D boundary Haag duality directly or prove the folding equivalence in enough detail to make the transfer legitimate.","section":"§4.4, boundary algebraic Haag duality paragraph"},{"comment":"The proof that the functor Y : Z_A(X) -> DHR^∂(X) is braided is only carried out in detail for A = Hilb, after which the general case is dismissed by saying that the A-strands going 'back into the page' play no role. This is not a proof: the A-enrichment interacts with the braiding through the half-braiding of A in Z(X), and the centralizer condition z ∈ Z_A(X) is exactly what would need to be used to verify that Y(β_{w,z}) agrees with the braiding u_{Λ,Δ}^{Y_w,Y_z} for general A. A complete verification is required before Theorem D and E can be accepted as braided equivalences.","section":"Theorem 4.21"},{"comment":"The braided tensor product X^{⊗Λ} over a rectangle is defined via a braid group action coming from choices of homeomorphisms, but the only cited reference, [SW03, Prop. 7.6], is stated in the balanced setting. The paper asserts that balancing is not needed. Since the net structure, including the inclusions X(Λ) ⊂ X(∆) used throughout Section 4 and in Definition 4.16, depends on this construction, the authors should either provide a complete proof for arbitrary braided fusion categories or give a precise reference that covers the non-balanced case.","section":"§4, Definition 4.3 and Remark 4.2"}],"minor_comments":[{"comment":"The statement of Theorem C should make explicit that the equivalence is only at the level of W*-categories; the footnote already says the monoidal and braided structures are not analyzed, but the theorem text reads as if the full braided tensor category is identified. Please align the theorem statement with the actual claim.","section":"Theorem C and footnote 1"},{"comment":"The sentence 'as we are ignoring the fusion and braiding, it suffices to consider the case of a 1D fusion spin system' needs a short justification, since the 2D braided categorical net is not literally a 1D fusion spin system even at the level of the underlying W*-category.","section":"§4.1, proof of Theorem 4.8"},{"comment":"These theorems are proved by saying that the arguments of [JNPW23, Lem. 4.7 and Thm. 4.8] can be adapted using a variant of the tube/sphere/dome algebra. Since these theorems identify the boundary algebras that are central to the paper, the authors should provide more details or give precise references to the specific statements being adapted.","section":"Theorems 3.8, 5.6, and 5.7"},{"comment":"There are several typographical and notation issues: 'consdider' in §4, inconsistent use of DHR^∂ vs DHR ∂, and the notation X^{⊗Λ} vs A^{⊗Λ} in Definition 4.16. A careful proofreading pass is recommended.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":"The paper is strong and the Levin-Wen half is well supported, but the folding-trick step in §4.4 is the single most important technical gap: without a proof that boundary DHR bimodules survive the dimensional reduction, Theorems D and E are not established. The braiding verification for general A in Theorem 4.21 also needs real work rather than a diagrammatic assertion. These issues are fixable within the paper's scope, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Two things to know. This is a genuine advance in the operator-algebraic approach to topological order, and its Levin-Wen analysis is on solid ground. But the main Walker-Wang boundary theorem hinges on an unproved folding-trick identification, so read §4.4 carefully before relying on Theorem D.\n\nThe paper extends the authors' LTO framework to topological boundaries: new boundary LTO axioms, a boundary DHR bimodule category, and explicit computations for Levin-Wen and Walker-Wang. Theorem A for Levin-Wen rests on established subfactor machinery and looks well supported. The construction of braided categorical nets is new and useful; Theorem B, showing type I cone von Neumann algebras, is a clean contrast with the type II/III cones in the earlier paper. Theorem C's W*-equivalence to Hilb(B) is stated honestly, with the fusion and braiding explicitly left open.\n\nThe soft spot is exactly what the stress-test note flags. In §4.4, the identification of boundary DHR bimodules of the 2D braided-enriched net X with those of the folded 1D net M is asserted in a sentence. No functor is constructed, and no argument shows localizability is preserved under the bijection. The subsequent chain DHR^∂(M) ≅ End(M_C) ≅ Z_A(X) therefore identifies the DHR category of the folded net, not necessarily of X. The braided equivalence in Theorem 4.21 inherits the gap: the proof is sketched for A = Hilb, and the general case is dismissed by saying the A-strands play no role. This is a localized proof gap, not a counterexample, but Theorem D is conditional as written. Theorem C also only identifies the W*-category; the conjectured fusion and braiding remain open, which the authors acknowledge.\n\nThe citation pattern is fine. The self-citations point to published, proven results, and there are no fitted parameters or hidden data. The paper reads as careful and intellectually honest.\n\nWho is this for? Anyone working in operator-algebraic approaches to topological order, boundary algebras, or holographic dualities. It deserves a serious referee, not a desk reject. My verdict matches the conditional: the Levin-Wen half is solid, the framework is plausible, and the folding step looks fixable, but the Walker-Wang boundary claims should not be accepted without that proof.","headline":"A genuine advance in the operator-algebraic approach to topological order, with a solid Levin-Wen analysis but a load-bearing unproved folding trick in the Walker-Wang boundary theorem.","tokens_in":40862,"tokens_out":2679,"would_cite":true,"duration_ms":27219,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["46L37","81T45","81T05"],"pacs":[],"model":"deepseek-v4-flash","headline":"Boundary DHR bimodules recover the Walker-Wang boundary topological order.","keywords":["local topological order","boundary algebras","DHR bimodules","Walker-Wang model","Levin-Wen model","braided categorical nets","enriched center","subfactor theory"],"falsifier":"Run the folding bijection on a concrete example, say $\\mathcal{A} = \\operatorname{Vec}(\\mathbb{Z}/p)$ and a faithful $\\mathcal{A}$-enriched category $\\mathcal{X}$, and check whether every boundary DHR bimodule of the 2D braided categorical net restricted to the complement of a boundary rectangle becomes, under folding, a bimodule localizable in a boundary interval of the 1D net. A single sector whose folded image fails localizability, or two distinct boundary sectors that fold to isomorphic bimodules, would falsify the identification $\\operatorname{DHR}_{\\partial}(\\mathcal{X}) \\cong \\mathcal{Z}_{\\mathcal{A}}(\\mathcal{X})$.","tokens_in":39683,"feed_emoji":"🕸️","tokens_out":9271,"duration_ms":83051,"temperature":0.7,"pith_summary":"This paper sets out to make topological holography precise for spin systems that have a physical topological boundary, not just a cut in the bulk. It introduces boundary local topological order axioms and shows that a bulk-boundary system has a boundary algebra whose boundary DHR bimodules—the operator-algebraic analogue of superselection sectors—recover the boundary excitations. For Levin-Wen models with a gapped boundary built from a module category $\\mathcal{M}_{\\mathcal{C}}$, the boundary DHR category is the dual category $\\operatorname{End}(\\mathcal{M}_{\\mathcal{C}})$. For Walker-Wang models built from a unitary braided fusion category $\\mathcal{B}$ with a $\\mathcal{B}$-enriched fusion category $\\mathcal{X}$, the boundary DHR category is the enriched center $\\mathcal{Z}_{\\mathcal{B}}(\\mathcal{X})$. If true, this gives a model-independent way to extract boundary topological order from local operator algebras.","feed_headline":"Boundary algebra recovers Walker-Wang topological order","feed_subtitle":"New boundary LTO axioms fold a 2D Walker-Wang boundary into a 1D net whose DHR bimodules are the enriched center.","key_machinery":"The load-bearing machinery is the boundary LTO axiom system, which assigns to each boundary rectangle a boundary algebra $B^{\\partial}(I)$ independent of the surrounding rectangles, plus the folding trick of Section 4.4. To compute $\\operatorname{DHR}_{\\partial}(\\mathcal{X})$, the paper folds the 2D braided-enriched categorical net $\\mathcal{X}$ on a lattice with boundary into a 1D fusion module categorical net $\\mathcal{M}$ whose multifusion category is $\\mathcal{C} = \\mathcal{X}^{\\mathrm{mp}}\\boxtimes_{\\mathcal{A}}\\mathcal{A}\\boxtimes_{\\mathcal{A}}\\mathcal{X}$. It then uses the subfactor embedding machinery for AF actions of unitary multifusion categories—Q-system realization and the embedding of $\\operatorname{End}(\\mathcal{M}_{\\mathcal{C}})$ into bimodules—to show the folded DHR category is $\\operatorname{End}(\\mathcal{M}_{\\mathcal{C}}) \\cong \\mathcal{Z}_{\\mathcal{A}}(\\mathcal{X})$. The braiding on $\\operatorname{DHR}_{\\partial}(\\mathcal{X})$ is then built from local projective bases, extending the construction for DHR bimodules.","core_discovery":"The central claim is that the physical boundary of a topologically ordered bulk-boundary system carries a net of operator algebras whose localized bimodules—the boundary DHR bimodules—form a unitary braided tensor category that is exactly the topological boundary order. Theorem D states this for Walker-Wang models: given a unitary braided fusion category $\\mathcal{B}$ and a $\\mathcal{B}$-enriched unitary fusion category $\\mathcal{X}$ (equivalently, a braided central functor $\\mathcal{B}\\to\\mathcal{Z}(\\mathcal{X})$), the Walker-Wang bulk-boundary system is a boundary LTO, its boundary algebra $\\mathcal{X}$ is a braided categorical net, and $\\operatorname{DHR}_{\\partial}(\\mathcal{X})$ is unitarily braided equivalent to $\\mathcal{Z}_{\\mathcal{B}}(\\mathcal{X})$, the Müger centralizer of $\\mathcal{B}$ inside $\\mathcal{Z}(\\mathcal{X})$. The same mechanism gives Theorem E, identifying the bulk Walker-Wang DHR category with the Müger center $\\mathcal{Z}_2(\\mathcal{B})$, and Theorem A, identifying the Levin-Wen boundary DHR bimodules with $\\operatorname{End}(\\mathcal{M}_{\\mathcal{C}})$.","pith_inferences":["If the folding trick is made fully rigorous, the same dimensional-reduction strategy should compute boundary DHR categories for more general domain walls between nontrivial phases, replacing $\\mathcal{A}$ by the braided fusion category of the wall.","The identification $\\operatorname{DHR}_{\\partial}(\\mathcal{X}) \\cong \\mathcal{Z}_{\\mathcal{A}}(\\mathcal{X})$ suggests a completeness statement the paper does not make: every $\\mathcal{A}$-enriched fusion category arises as the boundary DHR category of some Walker-Wang bulk-boundary system, making the enriched center a complete invariant of boundary sector categories.","The type I cone algebras with finite-dimensional centers raise the prospect that a modified superselection theory for non-factorial type I nets could make the braiding of the superselection category directly visible in the net, without passing through the relative tensor product of bimodules.","One could test the folding step numerically on small fusion categories by constructing the folded 1D net from the ladder category and checking equality of dimensions of intertwiners between boundary DHR bimodules against $\\mathcal{Z}_{\\mathcal{A}}(\\mathcal{X})$."],"forward_implications":["For every $\\mathcal{B}$-enriched fusion category $\\mathcal{X}$, the topological boundary order of the $\\mathcal{B}$ Walker-Wang model is the enriched center $\\mathcal{Z}_{\\mathcal{B}}(\\mathcal{X})$, so boundary DHR bimodules provide a model-independent parameterization of boundary excitations.","The bulk DHR category of a Walker-Wang model is only the Müger center $\\mathcal{Z}_2(\\mathcal{B})$; the braided fusion 2-category of the full (3+1)D topological order is not captured by DHR bimodules, and the paper states that a 2-categorical extension of DHR bimodules would be needed.","Levin-Wen boundaries from $\\mathcal{M}_{\\mathcal{C}}$ have boundary DHR category $\\operatorname{End}(\\mathcal{M}_{\\mathcal{C}})$, matching the classical subfactor description of fusion module spin chains.","The braided categorical net in its topological boundary state has type I cone von Neumann algebras with finite-dimensional centers, and its superselection sectors are $\\operatorname{Hilb}(\\mathcal{B})$; this is the first example of a type I net with nontrivial superselection sectors.","Truncating the Walker-Wang boundary algebra to a fusion spin chain recovers $\\mathcal{B}$ as a braided fusion category, from which $\\mathcal{Z}(\\operatorname{Mod}(\\mathcal{B}))$ reconstructs the full bulk topological order."],"supporting_citations":[{"why":"Supplies the LTO axioms, canonical state, and boundary net construction that this paper extends to topological boundaries.","marker":"[JNPW23]"},{"why":"Introduces DHR bimodules and their braiding for quasi-local algebras; the boundary DHR category is defined analogously.","marker":"[Jon24]"},{"why":"Provides the unitary multifusion category embedding results used in Appendix A to prove full faithfulness of the DHR bimodule functors.","marker":"[BCE+25]"},{"why":"Q-system completion technique that realizes $\\operatorname{End}(\\mathcal{M}_{\\mathcal{C}})$ as bimodules over the fusion module spin chain.","marker":"[CHPJP22]"},{"why":"Supplies the enriched string-net model formalism and projector/skein formulas used for the Walker-Wang and boundary models.","marker":"[GHK+24]"},{"why":"Gives the $\\mathcal{B}$-enriched UFC $\\mathcal{X}$ as topological boundary data for Walker-Wang models and the microscopic models.","marker":"[HBJP23]"},{"why":"Introduces the module category boundary Hamiltonians used to define Levin-Wen boundary projections.","marker":"[KK12]"},{"why":"Defines the Müger centralizer and center, supplying the identification of the enriched center with the centralizer.","marker":"[M¨ ug03]"}],"fun_headline_variants":["Boundary DHR bimodules equal Müger centralizer in Walker-Wang","Walker-Wang boundary algebra yields type I cones, unlike Levin-Wen","New boundary LTO axioms recover topological order from 1D net","DHR bimodules of boundary net are the topological boundary order","Cone algebras in Walker-Wang are type I, capturing full UBFC"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument rests on the folding-trick assertion in Section 4.4: every boundary DHR bimodule of the 2D net $\\mathcal{X}$ is claimed to become, after folding, a boundary DHR bimodule of the 1D fusion module categorical net $\\mathcal{M}$ with $\\mathcal{C} = \\mathcal{X}^{\\mathrm{mp}}\\boxtimes_{\\mathcal{A}}\\mathcal{A}\\boxtimes_{\\mathcal{A}}\\mathcal{X}$, and this correspondence is said to preserve localizability and to have no extra or missing sectors. That assertion is not proved in the paper, and it is exactly what converts $\\operatorname{DHR}_{\\partial}(\\mathcal{X})$ into the enriched center $\\mathcal{Z}_{\\mathcal{A}}(\\mathcal{X})$.","fun_headline_variants_meta":{"raw":{"variants":["Boundary DHR bimodules equal Müger centralizer in Walker-Wang","Walker-Wang boundary algebra yields type I cones, unlike Levin-Wen","New boundary LTO axioms recover topological order from 1D net","DHR bimodules of boundary net are the topological boundary order","Cone algebras in Walker-Wang are type I, capturing full UBFC"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000841,"raw_usage":{"total_tokens":3732,"prompt_tokens":1081,"completion_tokens":2651,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":697,"completion_tokens_details":{"reasoning_tokens":2553}},"tokens_in":697,"tokens_out":2651,"duration_ms":19782,"temperature":1.0,"reasoning_tokens":2553,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T18:22:42.623774+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the folding bijection on a concrete example, say $\\mathcal{A} = \\operatorname{Vec}(\\mathbb{Z}/p)$ and a faithful $\\mathcal{A}$-enriched category $\\mathcal{X}$, and check whether every boundary DHR bimodule of the 2D braided categorical net restricted to the complement of a boundary rectangle becomes, under folding, a bimodule localizable in a boundary interval of the 1D net. A single sector whose folded image fails localizability, or two distinct boundary sectors that fold to isomorphic bimodules, would falsify the identification $\\operatorname{DHR}_{\\partial}(\\mathcal{X}) \\cong \\mathcal{Z}_{\\mathcal{A}}(\\mathcal{X})$.","supporting_citations":[],"review_version":2}