{"id":"ecb97d7c-be25-4cc2-b52b-d6fbd916e634","arxiv_id":"2506.22025","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Twisting the toric code by a 2-cocycle confines anyons directionally, producing dipole and fractal excitations, size-dependent logical operators, and 3D generalizations to surface and X-cube codes.","lead":"A new family of quantum memory models twists the toric code's mathematical structure to confine particle-like errors in one direction, creating restricted-mobility 'dipole' excitations and fractal-shaped symmetries. The construction changes how many logical qubits the code stores, depending on the lattice size, and is extended to 3D surface and X-cube codes.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The exact fractal-symmetry claim in Sec. IV.B is the load-bearing unverified step; a small-system commutator check should settle it.","rationale":"I read the strongest claim as having two pillars. Pillar (i) is that Eq. (8) is a commuting, frustration-free stabilizer code with confined anyons, dipoles, and parity-dependent logical operators. Pillar (ii) is that twisting both plaquette and vertex terms produces exact Sierpinski-type symmetries and unpaired logical operators encoding classical bits. Pillar (i) appears sound: the projective-representation algebra gives commuting local terms, the parity relations come from products of stabilizers, and the survival of one |G|-dimensional logical cycle explains the stated 'two qubits' in the odd case. I found no internal inconsistency there. Pillar (ii), however, is the second independent central claim and is presented without the supporting calculation that would make it credible. The condition px >= 2(n-1), py = 2n is stated as a fact; the mutual commutation of all fractal operators is asserted with a brief overlap-counting remark; and the claim that boundary violations cancel under vertical periodic boundary conditions is not demonstrated. This is precisely the kind of delicate cancellation that can fail on small or periodic lattices. The reader's weakest assumption pointed to unproven commuting-projector and boundary properties; I partially agree, but I would locate the sharpest risk in this fractal-symmetry step, because it is the basis for the paper's headline unpaired-logical-operator result. Since the concern is absence of verification rather than a known contradiction, the appropriate disposition remains the reader's CONDITIONAL verdict, pending the exact small-system check described above.","tokens_in":14953,"tokens_out":23750,"duration_ms":272745,"concrete_test":"Using exact arithmetic in the 4-dimensional local space of Sec. II, implement H^{alpha,beta} on a torus with py=4, px=2 (the smallest allowed n=2 case) and construct the operators F_p^g and F_v^chi from the Sierpinski pattern displayed in Sec. IV.B. Directly evaluate [F_p^g, h] and [F_v^chi, h] for every local plaquette and vertex term h, and evaluate [F_p^g, F_v^chi] for all centers. If every commutator vanishes and the F operators act nontrivially on the common +1 eigenspace of H^{alpha,beta}, the exact-symmetry and classical-encoding claim is verified for that size; any nonzero commutator, or F acting as a product of Hamiltonian terms, refutes the claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing unverified step is the exact fractal-symmetry claim in Sec. IV.B. The paper asserts that bow-tie/Sierpinski operators F_p^g and F_v^chi become symmetries for px >= 2(n-1), py = 2n with vertical periodic boundary conditions, and that all such operators commute among themselves. No derivation of the size condition, no cancellation count for boundary violations, and no proof of [F,F']=0 is given; the only justification is the heuristic sentence that overlaps involve 'an even number of times' or 'operators with opposite cocycle that commute.' These operators are the entire basis for the headline result that unpaired logical operators encode a classical ground state with topological protection. If a single overlap yields an uncancelled cocycle phase, or if any plaquette/vertex term fails to commute on the stated sizes, the unpaired-logical-operator claim collapses. The basic Eq. (8) commuting-projector property and the parity-dependent logical analysis are less brittle: they follow from standard projective-representation algebra and Ref. [12], and the 'two-qubits' count is consistent with a single surviving |G|-dimensional logical cycle. Thus the reader's broad concern is sharpest at the fractal-symmetry claim.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a family of twisted toric-code Hamiltonians for the group Z2xZ2, built from projective regular representations. The central claim is that the Hamiltonian in Eq. (8) is a commuting, frustration-free stabilizer code in which vertical strings of X^alpha_g create anyonic excitations with energy proportional to string length, while the bound pair X^alpha_g tensor X^bar-alpha_g forms a deconfined dipole mobile only vertically. The paper further claims system-size-dependent logical operators, with parity of the number of vertical plaquettes sometimes eliminating one or both qubits; proposes boundary Hamiltonians with specific anyon-condensation patterns; gives a PEPS representation of the ground state; and extends the construction to doubly twisted models with Sierpinski-type fractal symmetries and unpaired logical operators, as well as to 3D surface-code and X-cube models with dipole-loop and dipole-planon excitations.","tokens_in":15193,"tokens_out":3582,"duration_ms":40274,"significance":"If the claimed properties hold, the paper offers a useful and conceptually interesting mechanism for engineering directional anyon confinement in stabilizer codes, with concrete proposals for dipole bound states, size-dependent logical spaces, and 3D generalizations. The use of projective representations and 2-cocycles is elegant, and the extension to fractal operators with classical-like logical encoding is potentially valuable for hybrid classical-quantum code ideas. The paper is also commendably explicit about the algebraic ingredients and includes a PEPS construction. However, several load-bearing statements are asserted rather than proved, most notably the commuting-projector property of the central Hamiltonians and the exact fractal-symmetry claim; these need to be substantiated before the main results can be considered established.","major_comments":[{"comment":"The Hamiltonian in Eq. (8) is asserted to be a commuting, frustration-free stabilizer code, but no proof or explicit check is given in the text; the only justification is a pointer to Ref. [12]. Since every later statement about logical operators, anyon excitations, and ground-state degeneracy assumes this property, the manuscript should either provide a self-contained proof of commutation and frustration-freeness, or state precisely which theorem or construction in Ref. [12] applies and why it covers this twisted Hamiltonian.","section":"Section III, Eq. (8)"},{"comment":"The boundary Hamiltonians are proposed without demonstrating that they commute with the bulk Hamiltonian terms or that they are gapped and realize exactly the listed anyon-condensation patterns. The condensation statements are presented as assertions, and no explicit commutation relations or gap argument is supplied. Since the boundary phase structure is one of the paper's advertised results, each proposed boundary term needs a verifiable check that it commutes with the bulk and that the claimed condensate follows from the local boundary algebra.","section":"Section III.C, Eqs. (9)-(12)"},{"comment":"The central claim that the fractal operators F_p^g and F_v^chi become exact Hamiltonian symmetries for system sizes with p_x >= 2(n-1), p_y = 2n and vertical periodic boundary conditions is not derived. No cancellation count for boundary violations is given, and the heuristic statement that overlaps involve 'an even number of times' or 'operators with opposite cocycle that commute' is insufficient. Because these operators are the entire basis for the unpaired-logical-operator result, the paper needs a rigorous derivation of the size condition, a proof that all F_p^g and F_v^chi commute with every Hamiltonian term, and an explicit check that [F_p^g, F_{p'}^h]=[F_v^chi,F_{v'}^sigma]=[F_p^g,F_v^chi]=0 on the stated lattices.","section":"Section IV.B"},{"comment":"The logical-operator counting is not proved. The text concludes that for an odd number of vertical plaquettes the only logical operators are \\bar Z1 and \\bar X1, and that for an even number there are exactly two pairs satisfying \\bar X1^odd \\bar X1^even = \\bar Z2 and \\bar Y2^odd \\bar Y2^even = \\bar Z1, but no complete argument rules out additional logical operators. Since the claimed system-size-dependent degeneracy is a main result, the paper should provide a stabilizer-dimension calculation or an equivalent exhaustive analysis of the operator algebra showing that no further independent logical operators exist.","section":"Section III, logical-operator analysis"}],"minor_comments":[{"comment":"There are numerous typos and grammatical slips, for example 'followoing', 'aparent', 'seemly', and 'They loose some quantumness'. The manuscript would benefit from a careful proofreading pass.","section":"Throughout"},{"comment":"The identity X^{\\bar\\alpha}_g X^\\alpha_g = Z_{\\hat g} is used heavily but is introduced without derivation; it should be proved explicitly from the definitions of the projective representations and the chosen cocycle.","section":"Section II, Eq. (5)"},{"comment":"The figures referenced in the text are not present in the submitted manuscript text; without the actual figures, several deformation arguments and excitation diagrams are very hard to follow. Please ensure that all figures are included and legible.","section":"Section III, Figs. 1-3"},{"comment":"The fractal operators F_p^g and F_v^chi are described pictorially but never defined by an explicit algebraic expression. A precise definition in terms of products of X^alpha_g and Z^beta_\\hat g operators on the lattice would remove ambiguity.","section":"Section IV.B"},{"comment":"Reference [12] is cited only as 'Nat. Commun. 15 (2024)' without a title or article number; please provide the full citation so that the delegated stabilizer-code construction can be located by readers.","section":"References"}],"recommendation":"major_revision","confidential_remarks":"The paper's central commuting-projector property is delegated to the author's own Ref. [12], and the fractal-symmetry claim in Sec. IV.B is asserted without proof. I recommend that the editor require a self-contained proof or an explicit small-lattice commutator check for the fractal symmetries before publication. The logical-operator counting also deserves a rigorous stabilizer-dimension argument. These are fixable within the manuscript's scope, so I do not recommend rejection, but the current level of rigor is not sufficient for acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First: this is a real construction with a genuinely new effect. The twisted toric code Hamiltonian (8) gives directional confinement of fluxes, deconfined dipoles from bound confined pairs, and logical operators whose existence depends on lattice parity. I traced the operator algebra in Section III; it is consistent. The decorated Wilson loops and the absence of any undecorated vertical X-loop logical operator follow from the projective representation.\n\nThe paper's clean conceptual contribution is in Section III.B: two mechanisms to unconfine the confined anyons—binding left- and right-handed projectively represented operators into dipoles, and decorating with Zs to make dyons. That is a nice idea, and it is carried through to the 3D extensions.\n\nSoft spots, in order. The commuting/frustration-free property of Eq. (8) is delegated to the author's own Ref. [12]. That is acceptable if Ref. [12] indeed proves it, but the paper should state explicitly what is proven there. The boundary Hamiltonians in Section III.C are proposed without gappedness proofs; the condensation tables are plausible but asserted. Moderate gap, not fatal.\n\nThe biggest concern is the fractal-symmetry claim in Section IV.B. The bow-tie/Sierpinski operators are stated to become exact symmetries for px >= 2(n-1), py = 2n with vertical PBC, and all such operators are asserted to commute. No derivation of the size condition or cancellation count is given. This claim is load-bearing for the unpaired-logical-operator and classical-encoding result. A single overlap with an uncancelled cocycle phase would break it. The stress-test note is right: a small-system commutator check, say py = 4 or 6, would settle it. My prior is that it works, because the cocycle phases are abelian and the boundary violations should cancel by a parity argument consistent with the rest of the paper, but 'should work' is not a proof.\n\nThe 3D section is more speculative, as the author acknowledges. That is acceptable for an exploratory paper.\n\nOverall, the central 2D construction is sound in outline and the logical operator analysis is careful. The fractal claim needs verification. The paper deserves a serious referee, with instructions to check the fractal commutator and to clarify the dependence on Ref. [12]. I would not desk-reject it, and I would cite it if I worked on topological codes or fracton models.","headline":"A promising twisted toric code construction with directional confinement and dipoles; the main risk is the unverified fractal-symmetry claim in Sec. IV.B.","tokens_in":15671,"tokens_out":2867,"would_cite":true,"duration_ms":28505,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A 2-cocycle twist of the toric code makes anyon energy depend on direction, turning confined excitations into mobile dipoles and making the code's logical qubits depend on lattice parity.","keywords":["toric code","anyon confinement","2-cocycle twist","projective representation","logical operators","stabilizer codes","fractal symmetries","fracton"],"falsifier":"Exact diagonalization of $H^{\\alpha}$ on small tori with odd and even numbers of vertical plaquettes (for example $3\\times 4$ and $4\\times 4$) would check whether the ground-state degeneracy and the logical algebra follow the paper's parity predictions: four states with only $\\bar Z_1,\\bar X_1$ for odd sizes, and the relations $\\bar X_1^{\\mathrm{odd}}\\bar X_1^{\\mathrm{even}}=\\bar Z_2$, $\\bar Y_2^{\\mathrm{odd}}\\bar Y_2^{\\mathrm{even}}=\\bar Z_1$ for even sizes. The claim is falsified if the degeneracy does not depend on parity or if those operators do not obey the stated relations.","tokens_in":14707,"feed_emoji":"🌀","tokens_out":13767,"duration_ms":124202,"temperature":0.7,"pith_summary":"This paper claims that replacing the ordinary $X_g$ operators on the horizontal edges of a $\\mathbb{Z}_2\\times\\mathbb{Z}_2$ toric code by a projective pair $X^{\\bar\\alpha}_g\\otimes X^{\\alpha}_g$ concentrates the energy cost of certain anyon strings along one direction: a vertical string of $X^{\\alpha}_g$ operators creates an excitation on every plaquette it passes, so the energy grows with string length. Binding a left-oriented and a right-oriented version of that string cancels the bulk cost and leaves a deconfined dipole that can move only vertically; decorating a vertical string with $Z$ operators makes a deconfined dyon-like excitation, but only when the number of vertical plaquettes is even. Because of that even/odd condition, the logical operators depend on the lattice width: odd sizes keep the usual two qubits, while even sizes leave only paired logical operators and effectively remove a qubit. Twisting both plaquette and vertex terms produces Sierpinski-triangle fractal operators that become exact symmetries for certain system sizes and give unpaired logical operators storing classical bits, not qubits. The same twist applied to 3D surface and X-cube models yields dipole-loop and dipole-planon excitations.","feed_headline":"Toric code twist confines anyons along one direction","feed_subtitle":"Odd or even lattice widths determine which logical qubits survive the twist.","key_machinery":"The engine is the regular projective representation $X^{\\alpha}_g$ of $G=\\mathbb{Z}_2\\times\\mathbb{Z}_2$, defined by $X^{\\alpha}_g|h\\rangle=\\alpha(g,h)|gh\\rangle$ with the 2-cocycle $\\alpha$ of Eq. (1), so that products pick up phases and distinct nontrivial elements anticommute. Its conjugate $X^{\\bar\\alpha}_g$ satisfies $X^{\\bar\\alpha}_g X^{\\alpha}_g=Z^{\\hat g}$ and $[X^{\\alpha}_g,X^{\\bar\\alpha}_h]=0$, the identities that make decorated Wilson loops and dipole bound states possible. Inserting these operators into the plaquette term makes a vertical string of $X^{\\alpha}_g$ violate a plaquette on every step, which is the directional confinement; the tensor product $X^{\\bar\\alpha}_g\\otimes X^{\\alpha}_g$ cancels those violations and leaves dipoles with vertical-only mobility. Parity-dependent products of plaquette terms over the torus convert this confinement into size-dependent logical operators.","core_discovery":"The paper's central claim is that the Hamiltonian $H^{\\alpha}$ of Eq. (8), built from the projective representation $X^{\\alpha}_g$ and its conjugate $X^{\\bar\\alpha}_g$ of $G=\\mathbb{Z}_2\\times\\mathbb{Z}_2$, is a commuting, frustration-free stabilizer code whose X-type fluxes are confined along the vertical direction while Z-type charges remain deconfined. The paper identifies two ways to partially lift the confinement: binding $X^{\\alpha}_g$ with $X^{\\bar\\alpha}_g$ creates a dipole that is free only in the vertical direction, and decorating a vertical $X^{\\alpha}_g$ string with $Z^{\\hat g}$ creates a deconfined dyon-like string when the vertical plaquette number is even. As a result the logical operators are parity-dependent: for an odd number of vertical plaquettes the code retains the usual two logical qubits, whereas for an even number the surviving operators satisfy $\\bar X_1^{\\mathrm{odd}}\\bar X_1^{\\mathrm{even}}=\\bar Z_2$ and $\\bar Y_2^{\\mathrm{odd}}\\bar Y_2^{\\mathrm{even}}=\\bar Z_1$, and some logical operators disappear. For the fully twisted Hamiltonian $H^{\\alpha,\\beta}$, Sierpinski-pattern combinations of $X^{\\alpha}_g$ and $Z^{\\beta}_{\\hat g}$ commute with the Hamiltonian for sizes with $p_x\\ge 2(n-1)$ and $p_y=2n$ under vertical periodic boundary conditions; since these fractal operators commute among themselves, they encode classical bits rather than qubits. The same twisting is extended to 3D, where it confines string-like excitations of the surface code and planons of the X-cube code, producing dipole-loop and dipole-planon excitations.","pith_inferences":["The parity dependence of the logical operators suggests a direct finite-size test that the paper does not propose: compute the ground-state degeneracy on tori with odd versus even vertical widths and check that the logical algebra changes accordingly.","The unpaired fractal symmetries of $H^{\\alpha,\\beta}$ hint at a classically decodable, topologically protected memory; one could test whether the code has growing distance on larger lattices by exhaustive decoding simulations, something the paper leaves open.","The same twisting mechanism could be applied to other finite abelian groups beyond $\\mathbb{Z}_2\\times\\mathbb{Z}_2$ and to lattices with defects; the paper notes that lattice defects permute confined and unconfined anyons but does not analyze how defects alter the parity-dependent logical structure."],"forward_implications":["For odd vertical plaquette number, $H^{\\alpha}$ reproduces the two-qubit logical space of the ordinary toric code; for even number, the code loses a logical qubit and the surviving pairs obey $\\bar X_1^{\\mathrm{odd}}\\bar X_1^{\\mathrm{even}}=\\bar Z_2$ and $\\bar Y_2^{\\mathrm{odd}}\\bar Y_2^{\\mathrm{even}}=\\bar Z_1$.","The dipole formed by $X^{\\bar\\alpha}_g\\otimes X^{\\alpha}_g$ is deconfined only in the vertical direction and braids trivially with $Z$-strings; splitting it horizontally by $Z^{\\hat g}$ destroys its vertical mobility.","In $H^{\\alpha,\\beta}$, the Sierpinski fractal operators are exact symmetries for $p_x\\ge 2(n-1)$, $p_y=2n$ with vertical periodic boundary conditions, and because they commute with each other they encode classical bits rather than a quantum code.","For general finite abelian groups, the logical operator content is governed by the slant-product subgroup $K_\\alpha$ and the irrep subset $I=\\mathrm{Im}(\\iota^\\alpha)$, so the code stores a $|G|$-dimensional and a $|K_\\alpha|$-dimensional qudit.","In 3D, the same twisting confines string-like excitations of the surface code and planons of the X-cube code, producing dipole-loop and dipole-planon excitations with hybrid mobility."],"supporting_citations":[{"why":"Supplies the Kitaev toric-code/quantum-double model that the twisted Hamiltonian deforms and whose vertex term is kept.","marker":"[1]"},{"why":"Provides the commuting stabilizer-code and ground-state construction for the twisted Hamiltonian; the paper's bulk analysis rests on this reference.","marker":"[12]"},{"why":"Provides the classification of gapped boundaries of quantum doubles used to determine which anyons can condense at the proposed boundary terms.","marker":"[9]"},{"why":"Introduces the notion of topological frustration used to interpret the system-size-dependent loss of logical operators.","marker":"[10]"},{"why":"Gives the Newman-Moore model whose symmetry-defect excitations are the analogue invoked for the Sierpinski fractal excitations.","marker":"[17]"},{"why":"States the no-go result for fractal logical operators in 2D quantum codes that the paper invokes to argue its fractal operators encode a classical subspace.","marker":"[18]"},{"why":"Supplies the gauging/matrix-product-state characterization of the subgroup $K_\\alpha$ and irrep subset $I$ used to summarize logical operators for general abelian groups.","marker":"[19]"}],"fun_headline_variants":["Confined anyons create dipoles in twisted toric codes","Twisted toric code yields direction-dependent anyon traps","Lattice parity decides logical operators in twisted toric code","Double twist brings fractal excitations to toric codes","Dipole excitations emerge from confined anyons in 3D toric codes"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The paper's analysis assumes that every twisted Hamiltonian is exactly solvable in the sense that all its local terms commute and share a ground state that violates none of them, and that the proposed boundary Hamiltonians have the claimed gap and condensation behavior; the bulk commutation is asserted and delegated to reference [12] rather than proved.","fun_headline_variants_meta":{"raw":{"variants":["Confined anyons create dipoles in twisted toric codes","Twisted toric code yields direction-dependent anyon traps","Lattice parity decides logical operators in twisted toric code","Double twist brings fractal excitations to toric codes","Dipole excitations emerge from confined anyons in 3D toric codes"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000221,"raw_usage":{"total_tokens":1505,"prompt_tokens":1056,"completion_tokens":449,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":672,"completion_tokens_details":{"reasoning_tokens":377}},"tokens_in":672,"tokens_out":449,"duration_ms":5179,"temperature":1.0,"reasoning_tokens":377,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T22:13:22.056308+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Exact diagonalization of $H^{\\alpha}$ on small tori with odd and even numbers of vertical plaquettes (for example $3\\times 4$ and $4\\times 4$) would check whether the ground-state degeneracy and the logical algebra follow the paper's parity predictions: four states with only $\\bar Z_1,\\bar X_1$ for odd sizes, and the relations $\\bar X_1^{\\mathrm{odd}}\\bar X_1^{\\mathrm{even}}=\\bar Z_2$, $\\bar Y_2^{\\mathrm{odd}}\\bar Y_2^{\\mathrm{even}}=\\bar Z_1$ for even sizes. The claim is falsified if the degeneracy does not depend on parity or if those operators do not obey the stated relations.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Introduces the notion of topological frustration used to interpret the system-size-dependent loss of logical operators."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the Newman-Moore model whose symmetry-defect excitations are the analogue invoked for the Sierpinski fractal excitations."},{"cited_title":"Schuch, I","cited_arxiv_id":null,"evidence_quote":"States the no-go result for fractal logical operators in 2D quantum codes that the paper invokes to argue its fractal operators encode a classical subspace."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the gauging/matrix-product-state characterization of the subgroup $K_\\alpha$ and irrep subset $I$ used to summarize logical operators for general abelian groups."}],"review_version":1}