{"id":"3abe6cb7-0395-42ea-a55e-1ef2fc30c893","arxiv_id":"2506.06577","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Angular k-uniformity is proposed as a geometric isometry criterion that generalizes hyperinvariant tensor network constructions from 2D tilings to regular hyperbolic honeycombs in arbitrary dimension.","lead":"This paper introduces a geometric criterion called angular k-uniformity for building holographic quantum error-correcting codes on curved higher-dimensional lattices. It aims to provide a general design principle for hyperinvariant tensor networks that can capture features such as nontrivial boundary correlations and state-dependent reconstruction.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The correlation-suppression claim of the general no-go theorem rests on an unproven DAG flow model; if that model is not faithful, the systematic classification of hyperinvariant codes in arbitrary dimension lacks a foundation.","rationale":"The paper's central programme claim is that angular k-uniformity gives a systematic, geometry-aware classification of hyperinvariant codes in arbitrary dimension. The load-bearing step that converts this from an ansatz into a classification is Theorem 1 plus the DAG analysis in Sec. IV B, which decide for each (lattice,k) whether the network is hyperinvariant and whether boundary correlations are nontrivial. That step is not derived: the DAG description is explicitly interpretive, and the correlation-suppression argument uses graph degree data rather than the isometric structure of the multi-tensor blocks. Without a proof, the 'Corr.' and hyperinvariance entries in Table II are unsupported. The proposed operator-pushing test would directly check whether the sink heuristic is faithful for the paper's own flagship example. I agree with the reader that this is the weakest load-bearing assumption. The rate inconsistency in Sec. IV F is a separate concrete error, but it is less central than the missing foundation of the no-go theorem; even if the rate formula were corrected, the classification would still need a proof. The verdict remains REJECT.","tokens_in":18408,"tokens_out":17904,"duration_ms":205151,"concrete_test":"Take the explicit {5,3,4} construction of Sec. IV E, with the angular 2-uniform X-I code Xi(6) on an octahedron as the vertex tensor and Hadamard gates as the edge tensor. Perform an exact operator-pushing calculation for one multi-tensor block: start with a single-site Z boundary operator on a sink vertex and push it one layer inward. The DAG/sink model of Sec. IV B predicts that for k=k_max=2 the operator is trapped in a local loop and never reaches the next layer; if the operator instead appears nontrivially on the outgoing legs of the block, the DAG model is unfaithful and the general no-go theorem is unsupported. The calculation uses only the stabilizer data in Figs. 15 and 19 and can be done by exact linear algebra.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central classification in Table II and the claim that angular k-uniformity enables systematic construction depend on Theorem 1 (Sec. IV B). That theorem's correlation claim is justified only by a DAG heuristic: a sink vertex with k_max inputs supposedly traps a single-site boundary operator because adjacent vertices have at most k_max-1 inputs and can treat the operator as an unknown input, preventing propagation into the bulk. This is asserted, not derived from the isometry conditions. Operator evolution in a tensor network is governed by the actual multi-tensor isometry blocks; a vertex's in-degree does not by itself determine whether an operator is correctable or remains local. The theorem is stated without proof, yet it is used to fill the 'Corr.' column of Table II and to conclude that maximal angular k-uniformity yields trivial boundary correlations. If the DAG picture is not faithful, the no-go theorem collapses and the framework no longer systematically identifies which honeycombs and k values support hyperinvariant codes. A secondary concrete inconsistency, the code rate rho=2 reported in Sec. IV F for an isometric encoding, reinforces that the quantitative scaffolding is unreliable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a geometric refinement of k-uniformity, called angular k-uniformity, and claims that it provides a systematic framework for identifying and constructing hyperinvariant holographic tensor networks and codes on regular hyperbolic honeycombs in arbitrary dimension. The framework is illustrated with explicit CSS code constructions of the X-I family on polytopes such as the octahedron, together with Hadamard edge tensors, and with transfer-matrix estimates of code rate and distance. The paper also presents a general no-go theorem (Theorem 1) intended to explain when maximal angular k-uniformity forces trivial boundary correlations, and it extends the discussion to multi-angular uniformity, uberholography, complementary recovery, and heterogeneous/qLEGO architectures.","tokens_in":18644,"tokens_out":8212,"duration_ms":80712,"significance":"If the central claims were established, the paper would supply a useful geometry-aware design principle for hyperinvariant tensor networks beyond the 2D Evenbly constructions, and the explicit X-I code family is a concrete and potentially reusable ingredient. The transfer-matrix rate/distance calculations are also a concrete starting point. However, the central no-go theorem is stated without proof, the multi-tensor isometry verifications are graphical rather than algebraic, and the reported code rate contains an internal inconsistency. The geometric notion of angular k-uniformity is suggestive, but the systematic classification and construction claims currently outrun the demonstrated results.","major_comments":[{"comment":"Theorem 1 is the foundation of the claimed systematic classification, but it is stated without proof and the supporting argument is a DAG heuristic: the text asserts that a sink vertex with k_max inputs traps a single-site boundary operator because neighboring vertices have at most k_max-1 inputs and can treat the operator as an unknown input. This is not derived from the multi-tensor isometry conditions, and operator propagation in a tensor network is controlled by the isometry blocks rather than by graph in-degree. Since the \"Corr.\" column of Table II is filled using this theorem, the central no-go claim is currently unsupported. In addition, for the {5,3,4} honeycomb the maximal angular k would be 3 (the facet size of the octahedral vertex figure), yet the paper analyzes only k=1,2, so the theorem is never actually confronted with the constructed examples.","section":"Sec. IV B, Theorem 1"},{"comment":"The multi-tensor isometry conditions for the X-I code constructions are said to be \"graphically shown\" via operator pushing, but no algebraic verification is presented. For a claim that a given tensor network block is an isometry, one must specify the block, compute the contraction of the block with its adjoint, and verify the expected identity; graphical pushing of X and Z operators is suggestive but does not by itself establish the multi-tensor isometry, especially because the edge tensors are Hadamard gates that mix the two bases. This verification is load-bearing: without it, the constructed codes cannot be certified as hyperinvariant.","section":"Sec. IV E, Figs. 19-20"},{"comment":"The reported code rate ρ=2 for the {5,3,4} honeycomb is impossible for an isometric bulk-to-boundary encoding: it would mean that the number of logical bulk qubits is asymptotically twice the number of boundary physical qubits, so no isometric embedding of the logical space into the boundary space exists. This indicates an error in the rate formula (5)-(7), most likely in the definition of N_bulk(n) or in the u vector and normalization, and it undermines the quantitative scaling claims in this section. The rate formula should be corrected and checked against the explicit Ξ(6) construction.","section":"Sec. IV F, Eq. (9)"},{"comment":"The abstract and introduction claim that angular k-uniformity enables \"systematic identification and construction\" of hyperinvariant codes on regular hyperbolic honeycombs in arbitrary dimension, but Appendix A classifies only compact regular honeycombs and omits the quasi-compact families described in Sec. III A. Moreover, many entries in Table II (e.g., the k=5,...,12 rows for {3,3,3,5}) are not accompanied by explicit vertex-code constructions; Sec. V B.1 concedes that the construction table is incomplete. The classification therefore currently exceeds what is demonstrated.","section":"Appendix A and Sec. V B.1"},{"comment":"Multi-angular k-uniformity is underspecified: the input region I_in is an arbitrary disjoint union of angularly connected subsets, but no bound is imposed on the total size |I_in|. For the map from H_{L∪I_in} to the complement to be an isometry, one needs |L| + |I_in| ≤ n - |I_in|, i.e., |I_in| ≤ (n-|L|)/2. Without such a bound, the definition can demand an isometry that is dimensionally impossible. This needs to be fixed before the uberholography analysis in this section can be evaluated.","section":"Sec. IV D, Definition 4"}],"minor_comments":[{"comment":"The paragraph beginning \"As discussed in Refs. [4–7], HaPPY codes prohibit...\" appears twice nearly verbatim; one copy should be removed.","section":"Sec. IV B"},{"comment":"The last sentence of item 5 is incomplete: \"...may broaden the applicability of angulark-uniformity in practical settings.ximate eror correcting like [19]\" should be completed or deleted.","section":"Sec. V B.5"},{"comment":"The vectors u and v(n) and the notation n, k, p are not all defined before Eq. (6); in particular, u is introduced only after Eq. (7). Please define all symbols before use.","section":"Sec. IV F, Eq. (6)"},{"comment":"The phrase \"not universally non-trivial\" is ambiguous; the surrounding text interprets it as \"trivial correlations\", but the theorem statement should say exactly what is claimed.","section":"Sec. IV B, Theorem 1"},{"comment":"The truncation symbols t, tr, rr are used without definitions, and the color scheme in Table II will be hard to read when printed in grayscale; consider labeling rows directly.","section":"Appendix B, Table III"}],"recommendation":"major_revision","confidential_remarks":"I am skeptical about the DAG flow model and would want to see a proof of Theorem 1 and a corrected rate formula before publication. The paper is a preprint-style draft with several unfinished parts; if the authors can supply the missing derivations, the framework could be publishable. Given the hep-th scope, the relation to AdS/CFT is mostly by analogy; the editor may want a referee with tensor-network expertise to assess the graphical isometry proofs."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: there is a real idea here, and parts of it are worth stealing, but the paper as written does not prove its own central claim. The angular k-uniformity definition is a sensible refinement, and the X-I code family gives explicit constructions that genuinely go beyond the 2D Evenbly examples. The no-go theorem behind the whole hierarchy, though, is a sketch, and the code-rate formula has a number that cannot be right for an isometric code.\n\nWhat is new and good: defining isometry conditions through the vertex figure, so that 'angular sectors' replace planar intervals, gives a workable language for higher-dimensional HTNs. The X-I codes are concrete CSS codes on centrosymmetric polytopes; Theorem 2 (full polytope symmetry) is believable by inspection. The graphical operator-pushing checks for the multi-tensor isometries are the usual standard in this subfield—Evenbly and Steinberg do the same—so they are acceptable evidence, not a black mark. The appendix tables give a useful catalog of known and claimed constructions across honeycombs.\n\nSoft spots: Theorem 1 is stated without proof, and the DAG heuristic does not carry the weight that the classification in Table II puts on it. In-degree of a sink vertex is not enough to conclude that a single-site boundary operator gets trapped; whether it propagates is decided by the actual isometry blocks, and the paper never derives the flow picture from those blocks. The stress-test note is right on that. Separately, Eq. (9) reports ρ=2 for {5,3,4}; if ρ is the bulk-to-boundary qubit ratio, an isometric embedding requires boundary legs to be at least as many in the limit, so ρ≤1. A value of 2 means either a wrong counting convention or a genuine violation of isometry—both need fixing. This is a concrete red flag, not a nitpick. To be fair, the paper admits many of these gaps in Sec. V B, but the abstract overstates completeness.\n\nWho this is for: holographic tensor network people looking for new construction principles. They would get ideas from the X-I codes and the vertex-figure viewpoint, but they should not quote the no-go theorem until it is actually proven.\n\nRecommendation: send it to a serious referee. The concept is new enough and the examples specific enough that it deserves referee time, even though I expect major revision before the central claims are established. I would not cite the no-go result, but I might cite the definition and the X-I family after cleanup.","headline":"Real idea in the angular k-uniformity definition and the X-I code family, but the central no-go theorem is a sketch and the rate formula has a red flag; worth engaging, not ready as-is.","tokens_in":19124,"tokens_out":4158,"would_cite":false,"duration_ms":43071,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A geometric criterion, angular k-uniformity, determines which rotationally symmetric tensors can seed hyperinvariant holographic codes on regular hyperbolic honeycombs in any dimension.","keywords":["angular k-uniformity","hyperinvariant tensor networks","holographic quantum error correction","hyperbolic honeycombs","CSS codes","complementary recovery","uberholography","multi-tensor isometries"],"falsifier":"Evaluate the boundary two-point correlation function $\\langle Z_i Z_j\\rangle$ on the explicit angular $1$-uniform $\\Xi(6)$-based code on $\\{5,3,4\\}$; the classification predicts nontrivial correlations for $k=1$ and forced triviality at maximal $k$, so a numerical or exact evaluation that finds zero at all separations in the $k=1$ case, or nonzero in the maximal-$k$ case, would falsify it.","tokens_in":18166,"feed_emoji":"🕸️","tokens_out":13186,"duration_ms":113754,"temperature":0.7,"pith_summary":"The paper introduces a geometric criterion, angular $k$-uniformity, for deciding which rotationally symmetric tensors can seed hyperinvariant holographic codes on regular hyperbolic honeycombs of any dimension. The criterion refines earlier uniformity notions by asking for isometric behaviour only within angular sectors of a tensor's vertex figure, and the paper shows that this sector-level condition is exactly what matches the multi-tensor isometry blocks used in hyperinvariant networks. With this tool it proves a general no-go theorem: when the uniformity level $k$ reaches the facet size of the vertex figure, the network cannot be hyperinvariant and boundary correlations are trivial, whereas the genuinely higher-dimensional case $k=1$ supports nontrivial correlations. The paper supplies explicit rotationally invariant CSS vertex codes at every angular uniformity level used on honeycombs such as $\\{5,3,4\\}$ and $\\{5,3,3,4\\}$, and it computes code rate and distance scaling. If correct, the framework turns the search for higher-dimensional holographic codes into a geometric classification problem, with concrete predictions for correlations, complementary recovery, uberholography, and modular extensions.","feed_headline":"Angular k-uniformity lifts holographic codes to higher dimensions","feed_subtitle":"The criterion fixes which vertex tensors yield hyperinvariant codes with nontrivial boundary correlations in 3D and 4D.","key_machinery":"The central object is the vertex figure representation, which replaces each vertex tensor by the $(d-1)$-dimensional polytope whose vertices are the tensor's physical indices, together with the geometric uniformity condition defined on it: angular $k$-uniformity requires every strongly angularly connected subset of $k$ indices—a subset contained in one facet of the vertex figure and connected there—to define an isometry to the complementary output, while no larger subset does. This condition carries the argument because it converts the multi-tensor isometry constraints of hyperinvariant networks into a statement about angular sectors of a polytope, making the no-go theorem, the correlation behaviour, and the residual regions of complementary recovery all computable from the Schläfli symbol of the honeycomb. The constructive side of the machinery is the family of $X$–$I$ CSS codes, defined on centrosymmetric polytopes by pairing each vertex with its antipode and taking weight-two $Z$ stabilizers on antipodal pairs plus weight-four $X$ stabilizers on pairs of pairs; these codes are rotationally invariant by construction and realize every angular uniformity level used in the paper's tables.","core_discovery":"The paper's central claim is that hyperinvariance in holographic tensor networks is governed by a geometric refinement of uniformity that lives on the vertex figure of the lattice: a rotationally invariant vertex tensor $A$ is angular $k$-uniform when every strongly angularly connected set of $k$ physical indices—indices lying together in one facet of the vertex figure and connected in that facet's 1-skeleton—defines an isometry from the logical input plus that set to the complementary output, and no larger set does. On a regular hyperbolic honeycomb this angular sector structure matches the multi-tensor blocks exactly, so the isometry level $k$ becomes a design parameter. The paper proves a general no-go theorem: if $k$ equals the full facet size, the network is not hyperinvariant and cannot have universally nontrivial two-point boundary correlations; on $\\{5,3,4\\}$, angular $1$-uniformity produces genuine higher-dimensional multi-tensor blocks with nontrivial correlations, while $k=2$ gives 2D-style blocks with correlations suppressed on common $\\{5,4\\}$ planes. It then shows that a stronger condition, multi-angular $k$-uniformity, allows hyperinvariance, nontrivial correlations, and uberholography to coexist, and it constructs explicit $X$–$I$ CSS codes on centrosymmetric polytopes (the $\\Xi(2m)$ family) that realize the required angular uniformity on vertex figures from $d=2$ through $d=4$.","pith_inferences":["If angular $k$-uniformity is the right organizing principle, the same sector-based isometry test could be applied to numerically optimized tensor networks, where the angular sectors are not fixed by lattice symmetry but learned; the paper itself lists variational constructions as an open direction.","The dimensional pattern in the paper's tables suggests a general tension: as the facet size of the vertex figure grows with dimension, the gap between the no-go level and $k=1$ widens, so higher-dimensional hyperinvariant codes may need multi-angular isometries to keep correlations nontrivial.","A straightforward extension is to test the same criterion on simplicial honeycombs, where the paper's angular-connectivity definition needs modification; one prediction is that a mixed multi-angular condition would be required to recover the correlation results proven for non-simplicial lattices.","The $X$–$I$ code family may be reusable outside holography as a systematic source of rotationally symmetric CSS codes with prescribed isometry sectors, which could feed modular quantum-error-correction constructions."],"forward_implications":["On any non-simplicial regular hyperbolic honeycomb, the angular $k$-uniformity level of the vertex tensor fixes whether the network is hyperinvariant: maximal facet-sized $k$ yields a no-go situation, while $k=1$ yields genuinely higher-dimensional multi-tensor blocks with nontrivial two-point correlations.","For the $\\{5,3,4\\}$ honeycomb, the angular $1$-uniform construction leaves a two-dimensional residual region in complementary recovery, whereas the angular $2$-uniform construction leaves only a one-dimensional residual curve; the paper derives this from sector geometry.","Multi-angular $k$-uniformity is the condition under which hyperinvariance, nontrivial boundary correlations, and uberholography can coexist, so it extends the design space beyond perfect-tensor codes.","The explicit $\\Xi(2m)$ CSS codes realize the required rotational invariance and angular uniformity on all regular vertex figures in two, three, and four dimensions, with code rate and distance scaling computed through recursively defined transfer matrices.","The framework extends to heterogeneous networks and qLEGO-style modular constructions, so at least one vertex-tensor class must be (multi-)angular $1$-uniform to support nontrivial correlations in square-faced tilings with mixed vertex degrees."],"supporting_citations":[{"why":"Supplies the perfect-tensor holographic code whose trivial correlations and uberholography the paper uses as the contrast case.","marker":"[3]"},{"why":"Introduced hyperinvariant tensor networks and the multi-tensor isometry conditions that angular k-uniformity generalizes.","marker":"[4]"},{"why":"Provides the 2D no-go theorem and heterogeneous-code analysis that Theorem 1 extends to higher dimensions.","marker":"[5]"},{"why":"Gives the planar 2-uniform code family and the square-seed code that inspires the X-I constructions.","marker":"[6]"},{"why":"Establishes hyperinvariant-code properties, approximate complementary recovery, and residual-region geometry used as 2D baselines.","marker":"[7]"},{"why":"Defines k-uniform states, the uniformity notion that angular k-uniformity refines.","marker":"[8]"},{"why":"Defines planar maximally entangled states, the planar antecedent of angular uniformity.","marker":"[10]"},{"why":"Defines planar k-uniform states, another direct antecedent used in the hierarchy.","marker":"[11]"},{"why":"Supplies the conformal-quasicrystal layer structure used for the code-rate and distance transfer-matrix calculations.","marker":"[13]"}],"fun_headline_variants":["Angular k-uniformity determines hyperinvariance in higher D","Higher-dimensional holography from angular sector isometries","Hyperinvariant codes beyond 2D via angular uniformity","Angular rule predicts hyperinvariance in hyperbolic honeycombs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument's load-bearing premise is that information can only flow one way through the network, with each vertex receiving at most $k$ wires, and that a vertex already receiving the maximum number of wires will trap a local boundary operator; the paper asserts this flow picture rather than deriving it from the isometry conditions.","fun_headline_variants_meta":{"raw":{"variants":["Angular k-uniformity determines hyperinvariance in higher D","Higher-dimensional holography from angular sector isometries","Hyperinvariant codes beyond 2D via angular uniformity","Angular rule predicts hyperinvariance in hyperbolic honeycombs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001257,"raw_usage":{"total_tokens":5233,"prompt_tokens":1109,"completion_tokens":4124,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":725,"completion_tokens_details":{"reasoning_tokens":4055}},"tokens_in":725,"tokens_out":4124,"duration_ms":29408,"temperature":1.0,"reasoning_tokens":4055,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-07T05:53:40.276467+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Evaluate the boundary two-point correlation function $\\langle Z_i Z_j\\rangle$ on the explicit angular $1$-uniform $\\Xi(6)$-based code on $\\{5,3,4\\}$; the classification predicts nontrivial correlations for $k=1$ and forced triviality at maximal $k$, so a numerical or exact evaluation that finds zero at all separations in the $k=1$ case, or nonzero in the maximal-$k$ case, would falsify it.","supporting_citations":[{"cited_title":"Holographic quantum error-correcting codes: toy models for the bulk/boundary correspon- dence.Journal of High Energy Physics, 2015:1–55, 06 2015","cited_arxiv_id":null,"evidence_quote":"Supplies the perfect-tensor holographic code whose trivial correlations and uberholography the paper uses as the contrast case."},{"cited_title":"1-isometry","cited_arxiv_id":null,"evidence_quote":"Introduced hyperinvariant tensor networks and the multi-tensor isometry conditions that angular k-uniformity generalizes."},{"cited_title":"Hyper-invariant tensor networks and holography.Phys","cited_arxiv_id":null,"evidence_quote":"Provides the 2D no-go theorem and heterogeneous-code analysis that Theorem 1 extends to higher dimensions."},{"cited_title":"Hyperin- variant multiscale entanglement renormalization ansatz: Approximate holographic error correction codes with power-law correlations.Phys","cited_arxiv_id":null,"evidence_quote":"Gives the planar 2-uniform code family and the square-seed code that inspires the X-I constructions."},{"cited_title":"Far from perfect: Quantum error correction with (hyperinvariant) evenbly codes, 07 2024","cited_arxiv_id":null,"evidence_quote":"Establishes hyperinvariant-code properties, approximate complementary recovery, and residual-region geometry used as 2D baselines."},{"cited_title":"Holographic codes from hyperinvariant tensor networks","cited_arxiv_id":null,"evidence_quote":"Defines k-uniform states, the uniformity notion that angular k-uniformity refines."},{"cited_title":"Planar maximally entangled states.Phys","cited_arxiv_id":null,"evidence_quote":"Defines planar k-uniform states, another direct antecedent used in the hierarchy."},{"cited_title":"Harris, Nathan A","cited_arxiv_id":null,"evidence_quote":"Supplies the conformal-quasicrystal layer structure used for the code-rate and distance transfer-matrix calculations."}],"review_version":1}