betaFeedback

WorksPhysics and Astronomy

Local gluing classification and bulk-locked clock rings for ZN\mathbb{Z}_N models on a Klein bottle

Abstract

We classify local reflection gluings of the ZN\mathbb{Z}_N plaquette model on lattice Klein bottles and solve the obstructed literal seam. Within the monomial far-corner ansatz, all commuting gluings form a family labelled by a unit uu of ZN\mathbb{Z}_N. They exist for every NN at even seam length and, at odd length, exactly when −1-1 is a square modulo NN. We determine the anyon transport, including exchange seams with P2=CP^2=C for charge conjugation CC, and the relation phases that distinguish an unfrustrated code from a commuting but frustrated Hamiltonian. For the literal gluing, the 2Lx2L_x noncommuting plaquettes form self-dual clock rings whose sectors are locked by the bulk. This gives an exact reduction of the spectrum to clock-ring spectra, closed-form ring ground energies for N=2,4N=2,4, and a closed-form ground-state degeneracy. We further determine twisted-coefficient homology on the stated surface domains and compare it with logical modules on registered lattice grids; composite dimensions need not give squares of cyclic groups. Dressed-isometry gates and monodromy are reported with their finite-grid scope. Finally, the qudit Gottesman–Knill theorem bounds the computational interpretation: Clifford deformations and Weyl measurements with stabilizer inputs and no external magic states cannot supply a non-stabilizer resource. Charge-conjugation measurements that do supply such states lie outside this operation model. The new results concern explicit seam classification, phase-sensitive consistency and bulk-locked spectra, not a new general no-go theorem or a universal computing architecture.

Introduction

A reflection gluing can preserve a lattice check algebra, frustrate its vacuum sector through a relation phase, or produce a noncommuting seam Hamiltonian. These possibilities are distinct even when the bulk topological order is unchanged. We study them explicitly for the ZN\mathbb{Z}_N plaquette model on a Klein bottle. Our main results are a classification of commuting local gluings within a stated ansatz and an exact reduction of the obstructed literal seam to bulk-locked clock rings.

In [1] we studied the Z2\mathbb{Z}_2 toric code on a lattice Klein bottle. In the edge register the reflection gluing preserves the checks; in the site register (Wen’s plaquette model [2]) a Hadamard dressing restores commutation, while the literal gluing produces locked Majorana rings. The extension to ZN\mathbb{Z}_N adds structures absent at N=2N=2: a unit-labelled uu-family, an arithmetic obstruction at odd circumference, order-four exchange transport with P2=CP^2=C, and logical modules with composite-NN torsion. Here CC is charge conjugation on anyon labels. None of these label relations alone specifies a measurement or a non-Clifford resource.

Contributions and results map. Theorems 1 and 2 classify all commuting gluings within the local monomial far-corner ansatz, determine their seam actions and code dimensions, and track the relation phases needed for an unfrustrated code. Theorems 3 and 4 reduce the non-commuting literal seam to bulk-locked ZN\mathbb{Z}_N clock rings with bulk-constrained sectors. These are the two central microscopic results. Theorem 5, on its stated domain, and the separate projective-plane calculation in Example 1 describe twisted homology. Propositions 2, 3 and 4 compare logical modules, dressed isometries and monodromy with their stated proof, argument and grid support. Finally, Corollary 1 and Remark 2 delimit the computation allowed by the operation model below; Proposition 5 identifies the non-stabilizer measurements in the Wigner examples.

Relation to known results. Non-orientable qudit homological codes and parity-invariant-anyon counts are known [3, 4], as are reflection-enriched phases [5] and cross-cap Clifford gates at N=2N=2 [6]. Recent quotient-code methods connect twisted compactification to logical operators and gates [7]. We do not claim these general principles, the mere possibility of order-four reflection transport, or stabilizer closure as new. Our contribution is their resolution in the specified plaquette realization: the complete local uu-gluing classification, its phase-sensitive consistency conditions, and the locked clock-ring spectrum of the literal obstruction. The composite-NN and gate results retain the evidence levels of Table 1. § 7 compares the constructions and assumptions in detail. The N=2N=2 results of Ref. [1] are re-derived rather than re-claimed.

resultcontentsupportprovenance
Thm. 1commuting gluings = uu-family; odd LxL_x iff u2≡−1u^2 \equiv-1proof; enumeration; gridpred. E-1 H1, H2
Thm. 2(i, iii)∣R∣|R|; frustration iff LxL_x odd and N≡2(mod4)N \equiv2 \pmod{4}; Lemma 1proof; gridpred. E-1 H3, H4; Lemma 1 post hoc
Thm. 2(ii, iii)seam action PP; Pe2=CP_e^2=C; ∣R∣=∣ker⁡(1−P)∣|R|=|\ker(1-P)|; phasesproof; sealed property checkscross (not pre-registered)
Thm. 3, 4literal gluing: clock rings, locking rule, spectrum, E0E_0, GSDproof; grid; numericspred. E-2 H1–H4 (reduced coverage)
Prop. 1row-exchanging reflection σ\sigma; Kramers–Wannier actionproof; gridpred. E-2 H5
Prop. 2logical module = twisted homology (edge / site register)proof / grid, argumentpred. E-3 H1
Sec. 5.1logical operators; distances (at the listed sizes)gridpred. E-3 H2, H3
Prop. 3admissible dressed isometries, logical actions, gate groupsargument; gridpred. E-3 H4, H5
Thm. 5Σg,k\Sigma_{g,k}: consistency, ∣H1∣|H_1|, block quotient, direct sumsproof; gridpred. E-4 H1; other structures: reading
Remark 1CC-defect fusion count consistent with ∣H1∣|H_1| (NN odd)arithmetic consistencypred. E-4 H2
Prop. 4double-winding closure counts; monodromy multiplier hhgridpred. E-4 H3; hh: reading
Cor. 1no non-stabilizer resource from cross-caps within OOproof (Gottesman–Knill)not computed
Prop. 5Wigner minima; outcome probabilities of three measurementsproof; gridpred. E-4 H4
Sec. 6.2no protection from non-orientability itselfhand assessmentnot computed

Table 1. Results and their evidence (labels as in Sec. 1; grids in Appendix A.1). “pred.”: pre-registered prediction (stage, hypothesis); “cross”: sealed cross-check. Check counts and details: Appendix A.3.

Operation model and computational boundary. For the computational discussion we specify O\mathcal{O}: static codes generated by ZN\mathbb{Z}_N Weyl stabilizers; code deformations implemented by Clifford unitaries and Weyl measurements; classical feed-forward; stabilizer inputs and no external non-stabilizer states. Lattice isometries with single-site Clifford dressings and cross-cap moves belong to O\mathcal{O} only when implemented by these operations. The qudit Gottesman–Knill theorem then excludes production of a non-stabilizer state, independently of the seam classification. In proposals using Z3\mathbb{Z}_3 topological order, additional resources enter through a symmetry-constrained charge measurement [8] or coherent interferometry with charge-conjugation defects [9]. Extra symmetry and fractionalization data, cross-cap-state preparation, coherent superpositions of mobile-defect paths, dynamical non-Pauli measurements and non-Clifford pumps are not supplied by our static gluing and lie outside O\mathcal{O}.

Scope. The classification is restricted to local monomial far-corner gluings of the fixed bulk plaquette model. We construct lattice codes only on the alternating Klein bottle; elsewhere we compute homology without asserting a lattice realization. In the exchange branch the double-winding external monodromy is P±2=CP^{\pm2}=C, but a single passage is PP, and at N=2N=2 the exchange e↔me\leftrightarrow m is not C=1C=1. Fusion counting is a consistency check, not an implementation of a defect theory. The no-go applies to O\mathcal{O}, not to arbitrary physical manipulations of cross-caps. No protected non-Clifford implementation, noise threshold or universal architecture is established.

Method and evidence. The work was done in five stages: E-0, an engine calibrated against the N=2N=2 results of Ref. [1] (no pre-registration), and E-1 to E-4, whose predictions and grids were frozen before the stage’s runs; results known beforehand are listed in each pre-registration and not counted as predictions. Each final run was sealed by sha256 after development or rehearsal runs (Appendix § 10). Table 1 labels each result by its support—proof, argument (a sketch whose general validity is not established), grid (exhaustive on the pre-registered grid), numerics—and its provenance: pre-registered prediction, cross-check (sealed, not pre-registered), reading (sealed, without registered expectation) or post hoc (after sealing, unsealed). Results of Ref. [1] that are the case N=2N=2 of ours are re-derived, not re-claimed.

Model and conventions

Weyl operators. With ω=e2πi/N\omega=e^{2\pi i/N}, X∣j⟩=∣j+1⟩X|j\rangle=|j+1\rangle and Z∣j⟩=ωj∣j⟩Z|j\rangle=\omega^j|j\rangle, so ZX=ωXZZX=\omega XZ. Weyl strings W(v)=∏sXsvsxZsvszW(v)=\prod_s X_s^{v_s^x}Z_s^{v_s^z}, v∈ZN2nv\in\mathbb{Z}_N^{2n} (up to phases, which all computations track exactly), satisfy W(v)W(w)=ω⟨v,w⟩W(w)W(v)W(v)W(w)=\omega^{\langle v,w\rangle}W(w)W(v) with ⟨v,w⟩=∑s(vszwsx−vsxwsz)\langle v,w\rangle=\sum_s(v_s^z w_s^x-v_s^x w_s^z).

Site register. A qudit sits on every site (x,y)(x,y), x∈ZLxx\in\mathbb{Z}_{L_x}, y=0,…,Ly−1y=0,\ldots,L_y-1. The plaquette p=(x,y)p=(x,y), with corners SW =(x,y)=(x,y), SE =(x+1,y)=(x+1,y), NE =(x+1,y+1)=(x+1,y+1), NW =(x,y+1)=(x,y+1), carries

Fp=XSWZSEXNE†ZNW†,H=−12∑p(Fp+Fp†),(1)F_p=X_{\mathrm{SW}}Z_{\mathrm{SE}}X_{\mathrm{NE}}^\dagger Z_{\mathrm{NW}}^\dagger,\qquad H=-\frac{1}{2}\sum_p(F_p+F_p^\dagger), \tag*{(1)}

the ZN\mathbb{Z}_N plaquette model [10]; for N=2N=2 it is Wen’s model, H=−∑pFpH=-\sum_pF_p [1, 2]. Plaquettes (x,y)(x,y) and sites (x,y)(x,y) have colour A (x+yx+y even) or B (odd). A site receives its XX-type factors from the two plaquettes of its own colour and its ZZ-type factors from the other two. Conjugating the odd sites by the Fourier transform F\mathcal{F} (FXF†=Z\mathcal{F}X\mathcal{F}^{\dagger}=Z, FZF†=X†\mathcal{F}Z\mathcal{F}^{\dagger}=X^\dagger) makes the A plaquettes XX-type and the B plaquettes ZZ-type: on a torus with even sides Eq. (1) is the ZN\mathbb{Z}_N toric code of the lattice whose vertices, edges and faces are the A plaquettes, the sites and the B plaquettes. A plaquette with eigenvalue ωk\omega^k carries charge kk; the charges of the two colours are the anyons ee and mm, with labels (A,B)∈ZN2(A,B)\in\mathbb{Z}_N^2, braiding phase ωAB′+A′B\omega^{A B'+A' B} and spin ωAB\omega^{A B}; a one-site translation exchanges them [2].

Edge register (control). Qudits sit on the oriented edges h(x,y):(x,y)→(x+1,y)h(x,y):(x,y)\to(x+1,y) and v(x,y):(x,y)→(x,y+1)v(x,y):(x,y)\to(x,y+1); AvA_v carries XX on the edges leaving vv and X†X^\dagger on those entering it, BpB_p carries Z±1Z^{\pm1} along the counterclockwise boundary (+1+1 along the arrow). This is the ZN\mathbb{Z}_N toric code [11], the homological code of the cellularization.

Mirror Klein bottle. Row LyL_y is identified with row 0 by x↦a−xx\mapsto a-x (offset aa). In the edge register the vertical edge leaving (x,Ly−1)(x,L_y-1) enters the vertex (a−x,0)(a-x,0), and the top edge of the plaquette (x,Ly−1)(x,L_y-1), traversed counterclockwise, is the row-0 edge h(a−x−1,0)h(a-x-1,0) traversed along its arrow, so the seam-row BpB_p carries ZZ there instead of Z†Z^\dagger. The code is again a homological code of a cellulation of the Klein bottle KK, and the seam reverses the sign of fluxes: P=diag⁡(1,−1)P=\operatorname{diag}(1,-1) on (e,m)(e,m). In the site register the seam plaquettes Sx:=F(x,Ly−1)S_x:=F_{(x,L_y-1)} have their far corners on row 0, NE at column j=a−x−1j=a-x-1 and NW at column j+1j+1 (Figure 1a); Rj:=F(j,0)R_j:=F_{(j,0)} are the row-0 plaquettes, all others are bulk plaquettes. There are Np=LxLy=nN_p=L_xL_y=n plaquettes and qudits. The reflection reverses the pattern X,Z,X†,Z†X,Z,X^\dagger,Z^\dagger, so the far corners must be chosen:

Figure 1. (a) Mirror Klein bottle of the site register (Lx=6L_x = 6, Ly=3L_y = 3, a=0a = 0; colours A/B). The virtual row LyL_y is row 0 read at column a−xa-x (top labels). A seam plaquette SxS_x keeps XX, ZZ on its near corners; its far corners land on row 0 at columns j=a−x−1j = a-x-1 and j+1j+1, where the uu-family places Z−ujZ^{-u_j} and X1/uj+1X^{1/u_{j+1}} (Theorem 1). (b) Non-commutation graph of the literal gluing (a=0a = 0): SxS_x (squares) and RjR_j (circles), labelled by column; SxS_x meets Ra−x−2R_{a-x-2} with phase ω\omega and Ra−xR_{a-x} with ω−1\omega^{-1}, giving two rings of length LxL_x (even LxL_x) or one ring of length 2Lx2L_x (odd LxL_x) (Theorem 3). For N=2N = 2 these are the graphs of Ref. [1], Fig. 2(a,b).

Definition 1 (local monomial far-corner gluing). Assign (pj,qj,pj′,qj′)∈ZN4(p_j,q_j,p'_j,q'_j)\in\mathbb{Z}_N^4 to every column jj of row 0. In SxS_x the NE far corner (column j=a−x−1j=a-x-1) carries X−pjZ−qjX^{-p_j}Z^{-q_j} and the NW far corner (column j+1j+1) carries X−pj+1′Z−qj+1′X^{-p'_{j+1}}Z^{-q'_{j+1}}; everything else is as in Eq. (1). The literal gluing has (pj,qj)=(1,0)(p_j,q_j)=(1,0), (pj′,qj′)=(0,1)(p'_j,q'_j)=(0,1): the far corners keep X†X^\dagger and Z†Z^\dagger.

Phases on the monomials affect only the scalars of relations Theorem 2(iii). By Theorem 1 the commuting gluings are the uu-gluings: the NE and NW far corners on column jj carry Z−ujZ^{-u_j} and X1/ujX^{1/u_j}, with uj=uu_j=u for even and −1/u-1/u for odd jj; u=±1u=\pm1 is the Fourier dressing (FF, F†F^\dagger alternating by column), and N=2N=2, u=1u=1 is the Hadamard-dressed gluing of Ref. [1]. For even LxL_x there are two branches. In the diagonal branch (a+Lya+L_y odd) every site still receives its XX-type factors from plaquettes of its own colour: the colouring continues across the seam, which maps each anyon type to itself. In the exchange branch (a+Lya+L_y even) the colours, and with them ee and mm, are exchanged across the seam. For odd LxL_x the colouring is inconsistent around xx and there is one charge type. “Grid” statements hold on the pre-registered grids (Appendix A.1) and are not extrapolated.

Local gluings: the uu-family

Theorem 1 (local gluings). Let N≥2N\ge2, Lx≥3L_x\ge3, Ly≥3L_y\ge3. A local monomial far-corner gluing makes all plaquettes commute if and only if, for every column jj,

pj=qj′=0,qjpj′≡−1,pj+1′≡qj(modN).(2)p_j=q'_j=0,\qquad q_jp'_j\equiv-1,\qquad p'_{j+1}\equiv q_j\pmod N. \tag*{(2)}

Equivalently, the NE and NW far corners on column jj carry Z−ujZ^{-u_j} and X1/ujX^{1/u_j} with uj∈ZN×u_j\in\mathbb{Z}_N^\times and uj+1=−1/uju_{j+1}=-1/u_j, so uj=uu_j=u for even and −1/u-1/u for odd jj. For even LxL_x every unit uu gives a commuting gluing (∣ZN×∣|\mathbb{Z}_N^\times| of them); for odd LxL_x the chain closes iff u2≡−1u^2\equiv-1. Hence commuting gluings with odd LxL_x exist iff −1-1 is a square mod NN, i.e. iff 4∤N4\nmid N and every odd prime factor of NN is ≡1(mod4)\equiv1\pmod4.

Proof. SxS_x shares far-corner sites only with Rj−1,Rj,Rj+1R_{j-1},R_j,R_{j+1} (j=a−x−1j=a-x-1; distinct as Lx≥3L_x\ge3) and with Sx±1S_{x\pm1}, whose far corners sit on columns j±1j\pm1; all other overlaps are those of the torus and cancel (Ly≥3L_y\ge3 keeps the row-0 plaquettes away from the seam row). Writing AB=ωe(A,B)BAAB=\omega^{e(A,B)}BA, e(Sx,Rj−1)=pje(S_x,R_{j-1})=p_j, e(Sx,Rj+1)=−qj+1′e(S_x,R_{j+1})=-q'_{j+1}, e(Sx,Rj)=pj+1′−qje(S_x,R_j)=p'_{j+1}-q_j, and e(Sx−1,Sx)=1+qj+1pj+1′−pj+1qj+1′e(S_{x-1},S_x)=1+q_{j+1}p'_{j+1}-p_{j+1}q'_{j+1} (one near and one far corner in common). Setting them to zero gives (2); then uj:=qju_j:=q_j is a unit and uj+1=−1/uju_{j+1}=-1/u_j. The recursion has period two, so it closes on even cycles for every uu and on odd ones iff u=−1/uu=-1/u. The last statement is the classical criterion for x2≡−1(modN)x^2\equiv-1\pmod N. □

For even LxL_x every unit works, and for odd LxL_x only the closing roots u2≡−1u^2\equiv-1; locally every unit is allowed because the uu-gluing is the quotient by the glide (x,y)↦(a−x,y+Ly)(x,y)\mapsto(a-x,y+L_y) dressed with X↦Z−usX\mapsto Z^{-u^s}, Z↦X1/usZ\mapsto X^{1/u^s} (us=uu_s=u or −1/u-1/u by sublattice), which maps every plaquette to some Fpu±1F_p^{u^{\pm1}} (derivation; checked post hoc): a symmetry of the stabilizer group for every uu, of HH only for u=±1u=\pm1. The arithmetic condition is the integer set of the time-reversal criterion for U(1)kU(1)_k Chern–Simons theory [12]; specific here is its role as a criterion for local lattice gluings.

Theorem 2 (seam action and code dimension). For a uu-gluing with Lx,Ly≥3L_x,L_y\ge3, let R={(κp)∈ZNNp:∏pFpκp∝1}R=\{(\kappa_p)\in\mathbb{Z}_N^{N_p}:\prod_p F_p^{\kappa_p}\propto1\} be the relation module and γ±=gcd⁡(u±1,N)\gamma_\pm=\gcd(u\pm1,N).

(i) ∣R∣=γ−γ+|R|=\gamma_-\gamma_+ in the diagonal branch, and ∣R∣=gcd⁡(2,N)|R|=\gcd(2,N) in the exchange branch and for odd LxL_x.

(ii) A charge carried upward through the seam has its label (A,B)(A,B) mapped, in the diagonal and the exchange branch respectively, by

Pd=(−u00u−1),Pe=(0−uu−10),(3)P_{\mathrm d}=\begin{pmatrix}-u&0\\0&u^{-1}\end{pmatrix},\qquad P_{\mathrm e}=\begin{pmatrix}0&-u\\u^{-1}&0\end{pmatrix}, \tag*{(3)}

and for odd LxL_x (one charge type) by −u-u with u2=−1u^2=-1. Both matrices invert the braiding phase and the spin, Pe2=−1=CP_{\mathrm e}^2=-1=C, and ∣R∣=∣ker⁡(1−P)∣|R|=|\ker(1-P)|, the number of PP-invariant labels, in every case.

(iii) With phase-free monomials all relation scalars are 11, except for odd LxL_x and N≡2(mod4)N\equiv2\pmod4, where (∏pFp)N/2=−1(\prod_pF_p)^{N/2}=-1 and no state satisfies all plaquettes. Phases change only the scalars: multiplying one seam plaquette by ω\omega leaves R\mathcal{R} unchanged and, for even NN in the exchange branch or with odd LxL_x, where R\mathcal{R} is generated by (∏pFp)N/2(\prod_p F_p)^{N/2}, flips its scalar. The code dimension is ∣R∣|\mathcal{R}| whenever all scalars are 1; only ∣R∣|\mathcal{R}| is an invariant of the gluing.

Lemma 1 (sign of the dressed product). For N≡2(mod4)N \equiv2 \pmod{4} and every uu-gluing, ∏pFpN/2=(−1)Lx\prod_p F_p^{N/2} = (-1)^{L_x}.

Proof. X~=XN/2\widetilde{X} = X^{N/2} and Z~=ZN/2\widetilde{Z} = Z^{N/2} satisfy X~2=Z~2=1\widetilde{X}^{2} = \widetilde{Z}^{2} = 1 and Z~X~=ωN2/4X~Z~=−X~Z~\widetilde{Z}\widetilde{X} = \omega^{N^2/4}\widetilde{X}\widetilde{Z} = -\widetilde{X}\widetilde{Z} (N/2N/2 odd), the qubit Pauli relations. FpN/2=X~SWZ~SEX~NEZ~NWF_p^{N/2} = \widetilde{X}_{\mathrm{SW}}\widetilde{Z}_{\mathrm{SE}}\widetilde{X}_{\mathrm{NE}}\widetilde{Z}_{\mathrm{NW}}, and as every uju_j is odd the far corners become Z~\widetilde{Z} (NE) and X~\widetilde{X} (NW): the product is that of the N=2N = 2 Hadamard-dressed gluing. Order its plaquettes by p=yLx+xp = yL_x + x and multiply site by site. A site with y≥1y \ge1 receives its four factors in the order XZZXXZZX (x≥1x \ge1) or ZXXZZXXZ (x=0x = 0): product +1+1. Row-0 site jj receives (Z,X)(Z, X) from Rj−1,RjR_{j-1}, R_j, reversed at j=0j = 0, followed by (Z,X)(Z, X) from the far corners of Sa−j−1,Sa−jS_{a-j-1}, S_{a-j}, reversed at j≡aj \equiv a; two pairs in the same order give XZXZXZXZ or ZXZX=−1ZXZX = -1, in opposite orders +1+1. So the sign is −1-1 at the Lx−2L_x - 2 sites j∉{0,a}j \notin\{0, a\} if a≢0a \not\equiv0, and at all LxL_x sites if a≡0a \equiv0; either way (−1)Lx(-1)^{L_x}.

For N=2N = 2 this sign was only observed, at the sizes computed, in Ref. [1]; the proof above was supplied by the internal review of draft v0.1 and checked post hoc.

Proof of Theorem 2. (i) ∏pFpκp∝1\prod_p F_p^{\kappa_p} \propto1 iff the exponents vanish on every site. A site with y≥1y \ge1 is SW and NE corner of (x,y)(x, y) and (x−1,y−1)(x - 1, y - 1) and SE and NW corner of (x−1,y)(x - 1, y) and (x,y−1)(x, y - 1), so κ\kappa is constant on colour classes, κA\kappa_A and κB\kappa_B (κA=κB\kappa_A = \kappa_B for odd LxL_x). Row-0 site jj gives κ(Rj)+uj−1κ(Sa−j)=0\kappa(R_j) + u_j^{-1}\kappa(S_{a-j}) = 0 and κ(Rj−1)−ujκ(Sa−j−1)=0\kappa(R_{j-1}) - u_j\kappa(S_{a-j-1}) = 0, and Sa−j,RjS_{a-j}, R_j have the same colour iff a+Lya+L_y is odd. Hence (u+1)κA=(u−1)κB=0(u+1)\kappa_A = (u-1)\kappa_B = 0 (diagonal), κB=uκA\kappa_B = u\kappa_A and 2κA=02\kappa_A = 0 (exchange), (u±1)κ=0(u \pm1)\kappa= 0 (odd LxL_x), with no division by non-units. (ii) (Derivation.) A ZqZ^q at the last site (x,Ly−1)(x, L_y - 1) of a string along (1,1)(1, 1) is continued through the NW far corner of SxS_x by Zqua−xZ^{q u_{a-x}} on row 0; an XqX^q string is continued through the NE far corner by Xq/ua−x−1X^{q/u_{a-x-1}}. Each string carries one colour, so a label continued on colour A is multiplied by −u-u and one continued on colour B by u−1u^{-1}, colours being preserved (diagonal) or exchanged (exchange branch): Eq. (3), up to a sign convention for labels. The entries of PP multiply to −1-1, which inverts AB′+A′BAB' + A'B and ABAB, and the equations of (i) are the fixed-point equations of PP. (iii) A relation supported on one colour has only XX-type or only ZZ-type factors on each site and scalar 1; in the diagonal branch these generate R\mathcal{R}. Otherwise, for even NN, R\mathcal{R} is generated by (∏pFp)N/2(\prod_p F_p)^{N/2}, whose scalar is 1 for N≡0(mod4)N \equiv0 \pmod{4} (XN/2X^{N/2} and ZN/2Z^{N/2} commute) and (−1)Lx(-1)^{L_x} for N≡2(mod4)N \equiv2 \pmod{4} (Lemma 1). Multiplying SxS_x by ω\omega multiplies that scalar by ωN/2=−1\omega^{N/2} = -1.

The diagonal branch with u=±1u = \pm1 is an order-two parity action, with ∣R∣=Ngcd⁡(2,N)|\mathcal{R}| = N\gcd(2,N) as for the edge register; the exchange branch realizes an order-four pair. Pe2=CP_e^2 = C (N>2N > 2), and encodes nothing for odd NN. For odd LxL_x the N=2N = 2 case is the sign-frustrated dressed gluing of Ref. [1]; on the grid the frustration recurs for N=10,26N = 10,26, and the codes are one-dimensional for N=5,13,17,25N = 5,13,17,25.

The obstructed seam: locked clock rings

Theorem 3 (obstruction of the literal gluing). Let N≥2N \ge2, Lx≥3L_x \ge3, Ly≥2L_y \ge2, any aa. In the literal gluing the non-commuting pairs are exactly {Sx,Ra−x−2\{S_x, R_{a-x-2} and {Sx,Ra−x}\{S_x, R_{a-x}\}, with SxRa−x−2=ωRa−x−2SxS_xR_{a-x-2} = \omega R_{a-x-2}S_x and SxRa−x=ω−1Ra−xSxS_xR_{a-x} = \omega^{-1}R_{a-x}S_x. They form two cycles of length LxL_x for even LxL_x and one of length 2Lx2L_x for odd LxL_x (Figure 1b). Along Sx→Ra−x−2→Sx+2→⋯S_x \to R_{a-x-2} \to S_{x+2} \to\cdots consecutive operators satisfy UiUi+1=ωUi+1UiU_iU_{i+1} = \omega U_{i+1}U_i with no operator inverted, so each cycle is a self-dual ZN\mathbb{Z}_N clock ring of m=Lx/2m=L_x/2 (even LxL_x) or m=Lxm=L_x (odd LxL_x) sites, with the seam plaquettes as field terms and the row-0 plaquettes as bond terms.

Proof. With j=a−x−1j=a-x-1, SxS_x carries X†X^\dagger on column jj and Z†Z^\dagger on j+1j+1 of row 0, and Rj′R_{j'} carries XX on j′j', ZZ on j′+1j'+1. For j′=j−1j'=j-1 the common site contributes +1+1, for j′=j+1j'=j+1 it contributes −1-1, and for j′=jj'=j both common sites carry commuting factors; other overlaps cancel as on the torus (also for Ly=2L_y=2). A step along the cycle advances the column by two, preserving its parity for even LxL_x and visiting all columns for odd LxL_x. U2i=XiU_{2i}=X_i, U2i+1=Zi†Zi+1U_{2i+1}=Z_i^\dagger Z_{i+1} represents the ring.

Let DrD_r and ErE_r be the products of the seam and of the row-0 plaquettes of ring rr (ring charge and ring twist); let Hclock(m;t,q)=−12∑i(Xi+Xi†)−12∑i(ωt[i=m−1]Zi†Zi+1+h.c.)H_{\mathrm{clock}}(m;t,q)=-\frac{1}{2}\sum_i(X_i+X_i^\dagger)-\frac{1}{2}\sum_i(\omega^{t[i=m-1]}Z_i^\dagger Z_{i+1}+\mathrm{h.c.}) on mm sites mod mm, restricted to ∏iXi=ωq\prod_i X_i=\omega^q, with lowest eigenvalue e0(m)e_0(m) at (t,q)=(0,0)(t,q)=(0,0), and let δ=2\delta=2 if LxL_x and a+Lya+L_y are both even, δ=1\delta=1 otherwise.

Theorem 4 (locking rule and exact reduction). For the literal gluing with N≥2N\geq2, Lx≥3L_x\geq3, Ly≥2L_y\geq2: (a) the relations form ZN\mathbb{Z}_N, generated by the product of all plaquettes and, for δ=2\delta=2, the product of the colour-A plaquettes, which commute pairwise; both scalars are 1, the product of all plaquettes taken in the order S0,…,SLx−1,R0,…,RLx−1S_0,\ldots,S_{L_x-1},R_0,\ldots,R_{L_x-1}, bulk (the index order gives the same); (b) the centre is generated by the bulk plaquettes FbF_b, the ring charges DrD_r and the ring twists ErE_r; with eigenvalues ωfb\omega^{f_b}, ωqr\omega^{q_r}, ωtr\omega^{t_r}, the realized joint eigenvalues are exactly those with

∑bfb+∑r(tr+qr)=0(modN)(4)\sum_b f_b+\sum_r(t_r+q_r)=0 \pmod N \tag*{(4)}

and, for δ=2\delta=2, ∑b∈Afb+qr+tr′=0\sum_{b\in A}f_b+q_r+t_{r'}=0, where rr is the ring whose seam plaquettes have colour A and r′r' the other; each realized joint eigenspace has dimension N2Lx−4+δN^{2L_x-4+\delta} (even LxL_x) or N2Lx−1N^{2L_x-1} (odd LxL_x), in which each ring carries one irreducible representation of dimension Nm−1N^{m-1} with multiplicity ν=NLx−2+δ\nu=N^{L_x-2+\delta} (even LxL_x) or NLxN^{L_x} (odd LxL_x), independent of LyL_y except through the parity of a+Lya+L_y; (c) every eigenvalue of HH is −∑bcos⁡(2πfb/N)+∑er∈spec⁡Hclock(mr;tr,qr)er-\sum_b\cos(2\pi f_b/N)+\sum_{e_r\in\operatorname{spec}H_{\mathrm{clock}}(m_r;t_r,q_r)}e_r, over the allowed labels, with multiplicity ν\nu times the ring multiplicities; (d) the ground states have all fb=0f_b=0 and every ring in (t,q)=(0,0)(t,q)=(0,0):

E0=−(Np−2Lx)+∑re0(mr),GSD=ν.(5)E_0=-(N_p-2L_x)+\sum_r e_0(m_r),\qquad\mathrm{GSD}=\nu. \tag*{(5)}

For N=2N=2 this is the result of [1]; for N=4N=4 a ring is two Ising chains and e0(m)=−2/sin⁡(π/2m)e_0(m)=-2/\sin(\pi/2m) as for N=2N=2, so every lattice has the same E0E_0 for N=4N=4 as for N=2N=2. For N=3N=3 and N≥5N\geq5 we give no closed form for e0(m)e_0(m) and use the numerical ground energy of the ring.

NNsizestatesE0E_0dev.GSD
24×34\times34 096−9.65685425-9.656854257×10−147\times10^{-14}8/168/16 d
34×24\times26 561−5.46410162-5.464101627×10−147\times10^{-14}81/2781/27 d
43×23\times24 096−4.00000000-4.000000005×10−145\times10^{-14}64/6464/64 d
33×33\times319 683−6.84232922-6.842329225×10−145\times10^{-14}27/2727/27 d, L
43×33\times3531 441−9.46410162-9.464101623×10−143\times10^{-14}27/8127/81 L
53×33\times31 953 125−7.23525950-7.235259501×10−141\times10^{-14}125/125125/125 L
35×35\times313 458 907−11.20032229-11.200322295×10−145\times10^{-14}243/243243/243 g
44×34\times316 777 216−9.65685425-9.656854256×10−146\times10^{-14}64/25664/256 g
63×33\times310 077 696−7.45670722-7.456707221×10−131\times10^{-13}216/216216/216 g

Table 2. Ground energies of the literal gluing, Eq. (5), against exact diagonalization on 9 of the 17 lattices of E-2 (dev. = largest deviation over a=0,1a=0,1). GSD (a=0/1a=0/1): counted from complete spectra (d: dense; b: block-resolved by bulk labels), or ν\nu of Theorem 4 in italics where Lanczos does not resolve degeneracies (L: with low levels; g: ground energy only).

Proof. (a) Each site carries one XX, X†X^\dagger, ZZ, Z†Z^\dagger, so relation coefficients are constant on colour classes; the seam identifies κ(Rj)\kappa(R_j) with κ(Sa−j−1)\kappa(S_{a-j-1}), which crosses colours unless a+Lya+L_y is even; nothing is divided, so composite NN adds no torsion. In the stated order every site contributes the scalar 1 except, for a≠0a\ne0, two row-0 sites contributing ω\omega and ω−1\omega^{-1}; the cancellation is global, not site by site. For δ=2\delta=2 the non-commuting pairs join opposite colours, and each site carries XX and X†X^\dagger, or ZZ and Z†Z^\dagger, from the colour-A plaquettes. (b), (c): the cycle algebra has a unique irreducible representation, of dimension Nm−1N^{m-1}, for each central character; count dimensions. For fixed tt, Hclock(m;t)H_{\mathrm{clock}}(m;t) on all NmN^m states has non-positive off-diagonal elements in the ZZ basis and is irreducible, so by Perron–Frobenius its ground state is unique, positive, and has q=0q=0. The ring translation exchanges DD and EE, so spec⁡(t,q)=spec⁡(q,t)\operatorname{spec}(t,q)=\operatorname{spec}(q,t) and min⁡spec⁡(t,0)>min⁡spec⁡(0,0)\min\operatorname{spec}(t,0)>\min\operatorname{spec}(0,0) for t≠0t\ne0. For N=4N=4, a Gray-code map writes the ring as two critical Ising chains.

:::

Proposition 1 (row-exchanging reflection). The site permutation σ:(x,0)↦(x,0)\sigma:(x,0)\mapsto(x,0), (x,y)↦(a−x,Ly−y)(x,y)\mapsto(a-x,L_y-y) for y≥1y\geq1, implemented by the unitary UσFjUσ†=Fσ(j)U_\sigma F_jU_\sigma^\dagger=F_{\sigma(j)}, UσFj†Uσ†=Fσ(j)†U_\sigma F_j^\dagger U_\sigma^\dagger=F_{\sigma(j)}^\dagger, is a symmetry of HH, and exchanges the seam row and row 0 (Rj↔Sa−j−1R_j\leftrightarrow S_{a-j-1} up to inverses), hence the field and bond terms of the rings: for odd LxL_x it acts on the ring as Ui↦Ui+Lx†U_i\mapsto U^\dagger_{i+L_x}, a Kramers–Wannier duality followed by a translation; for even LxL_x it exchanges the two rings. A uniform detuning of the seam row against row 0 breaks σ\sigma.

Proof. Lift to the double cover, where σ\sigma is y↦−yy\mapsto-y; bulk plaquettes map as (x,y)↦(a−x−1,Ly−1−y)(x,y)\mapsto(a-x-1,L_y-1-y).

For reweighted plaquette terms, invariance under σ\sigma and the half translation forces both ring masses to vanish for Lx≡2(mod4)L_x\equiv2\pmod4; for Lx≡0(mod4)L_x\equiv0\pmod4 a staggered @page 5/15 starts_mid=1 ends_mid=1 and {Sx,Ra−x}\{S_x, R_{a-x}\}, with SxRa−x−2=ωRa−x−2SxS_xR_{a-x-2} = \omega R_{a-x-2}S_x and SxRa−x=ω−1Ra−xSxS_xR_{a-x} = \omega^{-1}R_{a-x}S_x. They form two cycles of length LxL_x for even LxL_x and one of length 2Lx2L_x for odd LxL_x (Figure 1b). Along Sx→Ra−x−2→Sx+2→⋯S_x \to R_{a-x-2} \to S_{x+2} \to\cdots consecutive operators satisfy UiUi+1=ωUi+1UiU_iU_{i+1} = \omega U_{i+1}U_i with no operator inverted, so each cycle is a self-dual ZN\mathbb{Z}_N clock ring of m=Lx/2m=L_x/2 (even LxL_x) or m=Lxm=L_x (odd LxL_x) sites, with the seam plaquettes as field terms and the row-0 plaquettes as bond terms.

Proof. With j=a−x−1j=a-x-1, SxS_x carries X†X^\dagger on column jj and Z†Z^\dagger on j+1j+1 of row 0, and Rj′R_{j'} carries XX on j′j', ZZ on j′+1j'+1. For j′=j−1j'=j-1 the common site contributes +1+1, for j′=j+1j'=j+1 it contributes −1-1, and for j′=jj'=j both common sites carry commuting factors; other overlaps cancel as on the torus (also for Ly=2L_y=2). A step along the cycle advances the column by two, preserving its parity for even LxL_x and visiting all columns for odd LxL_x. U2i=XiU_{2i}=X_i, U2i+1=Zi†Zi+1U_{2i+1}=Z_i^\dagger Z_{i+1} represents the ring.

Let DrD_r and ErE_r be the products of the seam and of the row-0 plaquettes of ring rr (ring charge and ring twist); let Hclock(m;t,q)=−12∑i(Xi+Xi†)−12∑i(ωt[i=m−1]Zi†Zi+1+h.c.)H_{\mathrm{clock}}(m;t,q)=-\frac{1}{2}\sum_i(X_i+X_i^\dagger)-\frac{1}{2}\sum_i(\omega^{t[i=m-1]}Z_i^\dagger Z_{i+1}+\mathrm{h.c.}) on mm sites mod mm, restricted to ∏iXi=ωq\prod_i X_i=\omega^q, with lowest eigenvalue e0(m)e_0(m) at (t,q)=(0,0)(t,q)=(0,0), and let δ=2\delta=2 if LxL_x and a+Lya+L_y are both even, δ=1\delta=1 otherwise.

Theorem 4 (locking rule and exact reduction). For the literal gluing with N≥2N\geq2, Lx≥3L_x\geq3, $L_y\geq 2: (a) the relations form ZN\mathbb{Z}_N, generated by the product of all plaquettes and, for δ=2\delta=2, the product of the colour-A plaquettes, which commute pairwise; both scalars are 1, the product of all plaquettes taken in the order S0,…,SLx−1,R0,…,RLx−1S_0,\ldots,S_{L_x-1},R_0,\ldots,R_{L_x-1}, bulk (the index order gives the same); (b) the centre is generated by the bulk plaquettes FbF_b, the ring charges DrD_r and the ring twists ErE_r; with eigenvalues ωfb\omega^{f_b}, ωqr\omega^{q_r}, ωtr\omega^{t_r}, the realized joint eigenvalues are exactly those with

∑bfb+∑r(tr+qr)=0(modN)(6)\sum_b f_b+\sum_r(t_r+q_r)=0 \pmod N \tag*{(6)}

and, for δ=2\delta=2, ∑b∈Afb+qr+tr′=0\sum_{b\in A}f_b+q_r+t_{r'}=0, where rr is the ring whose seam plaquettes have colour A and r′r' the other; each realized joint eigenspace has dimension N2Lx−4+δN^{2L_x-4+\delta} (even LxL_x) or N2Lx−1N^{2L_x-1} (odd LxL_x), in which each ring carries one irreducible representation of dimension Nm−1N^{m-1} with multiplicity ν=NLx−2+δ\nu=N^{L_x-2+\delta} (even LxL_x) or NLxN^{L_x} (odd LxL_x), independent of LyL_y except through the parity of a+Lya+L_y; (c) every eigenvalue of HH is −∑bcos⁡(2πfb/N)+∑er∈spec⁡Hclock(mr;tr,qr)er-\sum_b\cos(2\pi f_b/N)+\sum_{e_r\in\operatorname{spec}H_{\mathrm{clock}}(m_r;t_r,q_r)}e_r, over the allowed labels, with multiplicity ν\nu times the ring multiplicities; (d) the ground states have all fb=0f_b=0 and every ring in (t,q)=(0,0)(t,q)=(0,0):

E0=−(Np−2Lx)+∑re0(mr),GSD=ν.(7)E_0=-(N_p-2L_x)+\sum_r e_0(m_r),\qquad\mathrm{GSD}=\nu. \tag*{(7)}

For N=2N=2 this is the result of [1]; for N=4N=4 a ring is two Ising chains and e0(m)=−2/sin⁡(π/2m)e_0(m)=-2/\sin(\pi/2m) as for N=2N=2, so every lattice has the same E0E_0 for N=4N=4 as for N=2N=2. For N=3N=3 and N≥5N\geq5 we give no closed form for e0(m)e_0(m) and use the numerical ground energy of the ring.

NNsizestatesE0E_0dev.GSD
24×34\times34 096−9.65685425-9.656854257×10−147\times10^{-14}8/168/16 d
34×24\times26 561−5.46410162-5.464101627×10−147\times10^{-14}81/2781/27 d
43×23\times24 096−4.00000000-4.000000005×10−145\times10^{-14}64/6464/64 d
33×33\times319 683−6.84232922-6.842329225×10−145\times10^{-14}27/2727/27 d, L
43×33\times3531 441−9.46410162-9.464101623×10−143\times10^{-14}27/8127/81 L
53×33\times31 953 125−7.23525950-7.235259501×10−141\times10^{-14}125/125125/125 L
35×35\times313 458 907−11.20032229-11.200322295×10−145\times10^{-14}243/243243/243 g
44×34\times316 777 216−9.65685425-9.656854256×10−146\times10^{-14}64/25664/256 g
63×33\times310 077 696−7.45670722-7.456707221×10−131\times10^{-13}216/216216/216 g

Table 2 (PDF p. 5). Ground energies of the literal gluing, Eq. (5), against exact diagonalization on 9 of the 17 lattices of E-2 (dev. = largest deviation over a=0,1a=0,1). GSD (a=0/1a=0/1): counted from complete spectra (d: dense; b: block-resolved by bulk labels), or ν\nu of Theorem 4 in italics where Lanczos does not resolve degeneracies (L: with low levels; g: ground energy only).

Proof. (a) Each site carries one XX, X†X^\dagger, ZZ, Z†Z^\dagger, so relation coefficients are constant on colour classes; the seam identifies κ(Rj)\kappa(R_j) with κ(Sa−j−1)\kappa(S_{a-j-1}), which crosses colours unless a+Lya+L_y is even; nothing is divided, so composite NN adds no torsion. In the stated order every site contributes the scalar 1 except, for a≠0a\ne0, two row-0 sites contributing ω\omega and ω−1\omega^{-1}; the cancellation is global, not site by site. For δ=2\delta=2 the non-commuting pairs join opposite colours, and each site carries XX and X†X^\dagger, or ZZ and Z†Z^\dagger, from the colour-A plaquettes. (b), (c): the cycle algebra has a unique irreducible representation, of dimension Nm−1N^{m-1}, for each central character; count dimensions. For fixed tt, Hclock(m;t)H_{\mathrm{clock}}(m;t) on all NmN^m states has non-positive off-diagonal elements in the ZZ basis and is irreducible, so by Perron–Frobenius its ground state is unique, positive, and has q=0q=0. The ring translation exchanges DD and EE, so spec⁡(t,q)=spec⁡(q,t)\operatorname{spec}(t,q)=\operatorname{spec}(q,t) and min⁡spec⁡(t,0)>min⁡spec⁡(0,0)\min\operatorname{spec}(t,0)>\min\operatorname{spec}(0,0) for t≠0t\ne0. For N=4N=4, a Gray-code map writes the ring as two critical Ising chains.

:::

Proposition 1 (row-exchanging reflection). The site permutation σ:(x,0)↦(x,0)\sigma:(x,0)\mapsto(x,0), (x,y)↦(a−x,Ly−y)(x,y)\mapsto(a-x,L_y-y) for y≥1y\geq1, implemented by the unitary UσFjUσ†=Fσ(j)U_\sigma F_jU_\sigma^\dagger=F_{\sigma(j)}, UσFj†Uσ†=Fσ(j)†U_\sigma F_j^\dagger U_\sigma^\dagger=F_{\sigma(j)}^\dagger, is a symmetry of HH, and exchanges the seam row and row 0 (Rj↔Sa−j−1R_j\leftrightarrow S_{a-j-1} up to inverses), hence the field and bond terms of the rings: for odd LxL_x it acts on the ring as Ui↦Ui+Lx†U_i\mapsto U^\dagger_{i+L_x}, a Kramers–Wannier duality followed by a translation; for even LxL_x it exchanges the two rings. A uniform detuning of the seam row against row 0 breaks σ\sigma.

Proof. Lift to the double cover, where σ\sigma is y↦−yy\mapsto-y; bulk plaquettes map as (x,y)↦(a−x−1,Ly−1−y)(x,y)\mapsto(a-x-1,L_y-1-y).

For reweighted plaquette terms, invariance under σ\sigma and the half translation forces both ring masses to vanish for Lx≡2(mod4)L_x\equiv2\pmod4; for Lx≡0(mod4)L_x\equiv0\pmod4 a staggered reweighting invariant under σ\sigma, translation and CC gives the rings opposite non-zero masses.

Self-dual clock rings are known: three-state Potts criticality for N=3N = 3, two Ising chains for N=4N = 4, and a critical phase containing the self-dual point for N≥5N \ge5 [13, 14]. As a pre-registered anchor, the N=3N = 3 ring gives the ratio of the lowest charged-sector gap (q=1q = 1) to the first neutral gap, Δσ/Δε=0.16200\Delta_{\sigma}/\Delta_{\varepsilon} = 0.16200 at m=12m = 12 (0.162980.16298 extrapolated in m−2m^{-2}; Potts ratio of scaling dimensions xσ/xε=1/6x_{\sigma}/x_{\varepsilon} = 1/6), and central charge c=0.7713c = 0.7713 from the registered fit, with the velocity 1.34781.3478 inferred from Δε\Delta_{\varepsilon} assuming xε=4/5x_{\varepsilon} = 4/5; both are within the registered ±10%\pm10\%, and the exact velocity 33/43\sqrt{3}/4 gives c=0.8002c = 0.8002 (reading). As for N=2N = 2, the ring degeneracy belongs to the frustrated Hamiltonian; we make no protection claim.

Arithmetic and codes

Klein-bottle codes

For a stabilizer group SS, S⊥S^{\perp} denotes the Weyl operators commuting with SS (modulo phases) and L=S⊥/S\mathcal{L} = S^{\perp}/S the logical module; the code dimension is ∣L∣1/2|\mathcal{L}|^{1/2}.

Proposition 2 (logical module). For the uu-gluing with even LxL_x, on the registered grid,

L≅H1(K;ZNρ)≅coker⁡(1+P)⊕ker⁡(1−P),(8)\mathcal{L} \cong H_1(K;\mathbb{Z}_N^{\rho}) \cong\operatorname{coker}(1 + P) \oplus\ker(1 - P), \tag*{(8)}

with ρ\rho trivial along rows and ρ=P\rho= P across the seam: (Zγ−⊕Zγ+)2(\mathbb{Z}_{\gamma_-} \oplus\mathbb{Z}_{\gamma_+})^2 in the diagonal branch and (Zgcd⁡(2,N))2(\mathbb{Z}_{\gcd(2,N)})^2 in the exchange branch. The edge register is the diagonal case u=−1u = -1, P=diag⁡(1,−1)P = \operatorname{diag}(1,-1), and L≅(ZN⊕Zgcd⁡(2,N))2\mathcal{L} \cong(\mathbb{Z}_N \oplus\mathbb{Z}_{\gcd(2,N)})^2 for every size.

Proof for the edge register; argument for the site register. In the edge register S⊥S^{\perp} consists of XX-type cocycles and ZZ-type cycles and SS of coboundaries and boundaries, so L≅H1(K;ZN)⊕H1(K;ZN)≅(ZN⊕Zgcd⁡(2,N))2\mathcal{L} \cong H^1(K;\mathbb{Z}_N) \oplus H_1(K;\mathbb{Z}_N) \cong(\mathbb{Z}_N \oplus\mathbb{Z}_{\gcd(2,N)})^2. In the site register, string operators modulo stabilizers are 1-chains whose label coefficients are twisted by PP across the seam; for π1(K)=⟨α,β∣βαβ−1α⟩\pi_1(K) = \langle\alpha,\beta\mid\beta\alpha\beta^{-1}\alpha\rangle, ρ(α)=1\rho(\alpha) = 1, ρ(β)=P\rho(\beta) = P, Fox calculus gives ∂2=(1+P,0)\partial_2 = (1 + P,0), ∂1=(0,P−1)\partial_1 = (0,P - 1), hence Eq. (8). In the diagonal branch, conjugating the odd sites by F\mathcal{F} gives a CSS code with the A and B plaquettes as checks and unit coefficients, i.e. H1⊕H1H^1 \oplus H_1 of a cellulation of KK with unit-twisted incidence numbers; that this complex (or its exchange-branch analogue) is the twisted complex of Eq. (8) is not proved here.

For composite NN, L\mathcal{L} is in general not the square of a cyclic group (Table 3): at N=8N = 8, u=±1u = \pm1 give (Z2⊕Z8)2(\mathbb{Z}_2 \oplus\mathbb{Z}_8)^2 and u=±3u = \pm3 give (Z2⊕Z4)2(\mathbb{Z}_2 \oplus\mathbb{Z}_4)^2; at N=12N = 12 all four units give (Z2⊕Z12)2(\mathbb{Z}_2 \oplus\mathbb{Z}_{12})^2, split differently between the colours. (At N=15N = 15, u=4u = 4 gives (γ−,γ+)=(3,5)(\gamma_-,\gamma_+) = (3,5) and (Z3⊕Z5)2≅Z152(\mathbb{Z}_3 \oplus\mathbb{Z}_5)^2 \cong\mathbb{Z}_{15}^2.)

NNuuγ−\gamma_-γ+\gamma_+L\mathcal{L}dim∣Γ∣|\Gamma|ℓ≠0\ell\ne0
2122Z22\mathbb{Z}_2^244yes
3133Z32\mathbb{Z}_3^232no
4142(Z2⊕Z4)2(\mathbb{Z}_2 \oplus\mathbb{Z}_4)^284yes
5211Z52\mathbb{Z}_5^211no
6162(Z2⊕Z6)2(\mathbb{Z}_2 \oplus\mathbb{Z}_6)^2124yes
8182(Z2⊕Z8)2(\mathbb{Z}_2 \oplus\mathbb{Z}_8)^2168yes
8324(Z2⊕Z4)2(\mathbb{Z}_2 \oplus\mathbb{Z}_4)^282no
10322Z24\mathbb{Z}_2^444yes
12546(Z2⊕Z12)2(\mathbb{Z}_2 \oplus\mathbb{Z}_{12})^2248yes
15435Z152\mathbb{Z}_{15}^215——
161162(Z2⊕Z16)2(\mathbb{Z}_2 \oplus\mathbb{Z}_{16})^23216yes
16728(Z2⊕Z8)2(\mathbb{Z}_2 \oplus\mathbb{Z}_8)^2164no

Table 3. Site codes at 4×34 \times3, a=0a = 0 (diagonal branch; E-3): logical module L\mathcal{L}, dimension ∣L∣1/2|\mathcal{L}|^{1/2}, order ∣Γ∣|\Gamma| of the group generated by the logical actions of all dressed isometries modulo logical Weyl operators (N=15N = 15 is not on the isometry grid), and whether the Dehn twist acts (ℓ≠0\ell\ne0). Edge register, 4×34 \times3, a=0a = 0: ∣Γ∣=2,2,4,4,4,6,8,8,10,8|\Gamma| = 2, 2, 4, 4, 4, 6, 8, 8, 10, 8 for N=2,…,12N = 2,\ldots,12.

Logical operators (grid). The row loop Λy(p,q)\Lambda_y(p,q) places XpZqX^pZ^q on the even and XqZpX^qZ^p on the odd sites of row yy; its label is qq on colour yy and pp on colour y+1y + 1. In the diagonal branch it commutes with SS, lies in SS iff its label (A,B)(A,B) has γ−∣A\gamma_- \mid A, γ+∣B\gamma_+ \mid B, and Λy(p,q)≡Λy+1(p,q)\Lambda_y(p,q) \equiv\Lambda_{y+1}(p,q) mod SS; hA=[Λ1(1,0)]h_A = [\Lambda_1(1,0)] and hB=[Λ1(0,1)]h_B = [\Lambda_1(0,1)] have orders γ−\gamma_-, γ+\gamma_+. The axis diagonals ΠyZ(x0+y,y)q\Pi_y Z^{q}_{(x_0+y,y)}, 2x0≡a−Ly+12x_0 \equiv a - L_y + 1, and ΠyX(x1−y,y)p\Pi_y X^p_{(x_1-y,y)}, 2x1≡a+Ly−12x_1 \equiv a + L_y - 1, close iff (1+ux0)q≡0(1 + u_{x_0})q \equiv0, resp. (ux1−1)p≡0(u_{x_1} - 1)p \equiv0, and are then non-trivial modulo SS and the row loops. Each colour has a one-site-per-row vertical representative, with two turns where the straight diagonals all carry the other colour (Lx/2L_x/2, LyL_y even; 74 codes). In the exchange branch with NN even, Λy(p,q)∈S\Lambda_y(p,q) \in S iff p+qp + q is even, and for even aa the colour sums Πy(XN/2ZN/2)(c,y)\Pi_y(X^{N/2}Z^{N/2})_{(c,y)}, c∈{a/2,a/2+Lx/2}c \in\{a/2,a/2 + L_x/2\}, are non-trivial and anticommute with Λy(1,0)\Lambda_y(1,0). With the four edge-register loop types, these and SS generate S⊥S^{\perp} on every grid code.

Distances (grid only; site register with u=±1u = \pm1, N≤8N \le8, and N=8N = 8, u=3u = 3 in the diagonal branch, u=1u = 1 and even NN in the exchange branch). Diagonal branch (dimension >1> 1): d=min⁡(Lx,Ly)d = \min(L_x,L_y); exchange branch: min⁡(Lx,2⌊Ly/2⌋)\min(L_x,2\lfloor L_y/2\rfloor); edge register, NN even: dX=min⁡(Lx,Ly+[a even])d_X = \min(L_x,L_y + [a\text{ even}]), dZ=min⁡(Lx,Ly+[a odd])d_Z = \min(L_x,L_y + [a\text{ odd}]) as for N=2N = 2; NN odd: dZ=Ly+[a odd]d_Z = L_y + [a\text{ odd}] and dX=Lxd_X = L_x at the computed sizes. The last fails off the grid: an XX-type logical made of two dual cross-cap cores and a seam-row string has weight 13 at N=3N = 3, 55, 14×314 \times3, and 14 at N=3N = 3, 16×316 \times3 (a=1a = 1; readings).

Lattice isometries and gates

The isometries of the lattice Klein bottle are the maps (x,y)↦(sxx+c,syy+cy)(x,y) \mapsto(s_xx + c,s_yy + c_y), sx,sy=±1s_x,s_y = \pm1, that normalize the deck group generated by (x,y)↦(x+Lx,y)(x,y) \mapsto(x + L_x,y) and (x,y)↦(a−x,y+Ly)(x,y) \mapsto(a - x,y + L_y): c∈{0,Lx/2}c \in\{0,L_x/2\} for sx=1s_x = 1, $c \in {a,a+Lx/2}\{a, a + L_x/2\} for sx=−1s_x = -1, cy∈ZLyc_y \in\mathbb{Z}_{L_y}; 8Ly8L_y maps in all [1]. The twist τ=1\tau= 1 iff cc is the half-period choice (the half translation is not isotopic to the identity and acts on homology as a Dehn twist [1]); the colour parity s=c+cy−[sx=−1]−[sy=−1] mod 2s = c + c_y - [s_x = -1] - [s_y = -1] \bmod2 is the change of plaquette colour under the map. A dressed automorphism is the induced site permutation followed by single-site Cliffords mapping each plaquette to a unit power of the plaquette at the image position; such dressings are monomial. The multiplier MλM_\lambda (λ∈ZN×\lambda\in\mathbb{Z}_N^\times) acts as X↦XλX \mapsto X^\lambda, Z↦Z1/λZ \mapsto Z^{1/\lambda} on even and with λ↔1/λ\lambda\leftrightarrow1/\lambda on odd sites. In the diagonal branch we use the basis (hA,hB,vA,vB)(h_A,h_B,v_A,v_B) of L\mathcal{L}: the row loops above and the classes of one-site-per-row vertical loops of colour A and B whose labels ℓA,ℓB\ell_A,\ell_B generate the two components of ker⁡(1−P)\ker(1-P) (orders γ+,γ−\gamma_+,\gamma_-); the edge register uses its ee and mm loops in the same roles. Modulo logical Weyl operators MλM_\lambda acts as diag⁡(λ−1,λ,λ−1,λ)\operatorname{diag}(\lambda^{-1},\lambda,\lambda^{-1},\lambda). Let ι:ker⁡(1−P)→coker⁡(1+P)\iota:\ker(1-P)\to\operatorname{coker}(1+P) be the natural map, sending a vertical loop to the row loop with the same label (vA↦(ℓA mod γ−)hAv_A\mapsto(\ell_A\bmod\gamma_-)h_A, vB↦(ℓB mod γ+)hBv_B\mapsto(\ell_B\bmod\gamma_+)h_B; 2ι=02\iota=0), and J=1+ιJ=1+\iota.

Proposition 3 (gates from lattice isometries). On the isometry grid, within the single-site dressing class: (a) A dressed automorphism exists iff, in the diagonal branch, (s=0 and (sy=1 or u2≡1))(s=0\text{ and }(s_y=1\text{ or }u^2\equiv1)) or (s=1 and ((sy=1 and u2≡−1) or N=2))(s=1\text{ and }((s_y=1\text{ and }u^2\equiv-1)\text{ or }N=2)); in the exchange branch, iff c+cy+[sy=−1]c+c_y+[s_y=-1] is even or −1-1 is a square mod NN; always in the edge register. The dressings of the identity are the multipliers (λ∈ZN×\lambda\in\mathbb{Z}_N^\times; λ2≡1\lambda^2\equiv1 in the exchange branch), and each admissible isometry has as many. (b) Modulo logical Weyl operators, colour-preserving dressed automorphisms in the diagonal branch and all those of the edge register act as MλJM_\lambda J; on codes of dimension >1>1 a non-trivial colour exchange occurs in the diagonal branch only for u2≡−1u^2\equiv-1 (on the grid: N=2N=2, and N=10N=10 with u=3,7u=3,7) and acts outside {Mλ}⟨J⟩\{M_\lambda\}\langle J\rangle; in the exchange branch (NN even) every dressed automorphism, including the geometric colour exchanges allowed by (a), is a logical Weyl operator. (c) The generated group has order ∣Zℓ×∣2[ι≠0]2[colour exchange]|\mathbb{Z}_\ell^\times|2^{[\iota\ne0]}2^{[\text{colour exchange}]}, ℓ=lcm⁡(γ−,γ+)\ell=\operatorname{lcm}(\gamma_-,\gamma_+) (diagonal), and ∣ZN×∣2[N even]|\mathbb{Z}_N^\times|2^{[N\text{ even}]} (edge), and ι≠0\iota\ne0 iff γ−γ+∤N\gamma_-\gamma_+\nmid N, i.e. iff NN is even and u≡±1(mod2ν)u\equiv\pm1\pmod{2^\nu}, 2ν2^\nu being the largest power of 22 dividing NN.

Necessity in (a) is an argument valid beyond the grid: the induced (anti-)automorphism χ\chi of the labels must satisfy χPχ−1=Psy\chi P\chi^{-1}=P^{s_y}; sufficiency, (b) and (c) are exhaustive on the grid. At N=2N=2 the isometries give SWAP⁡T(H⊗H)s\operatorname{SWAP}^T(H\otimes H)^s [1]; for ZN\mathbb{Z}_N the ingredients separate. The Dehn twist acts as v↦v+ι(v)v\mapsto v+\iota(v), a CNOT-type map on the order-two parts; for odd NN the half translation is a multiplier up to logical Weyl operators. A non-trivial colour exchange (u2≡−1u^2\equiv-1, dimension >1>1) forces N≡2(mod4)N\equiv2\pmod4 and dimension 44 for N>2N>2. The multipliers exist equally on the torus, where every isometry, colour exchanges included, has a dressing (controls at 4×44\times4, 6×66\times6, N≤8N\le8); within this class the scarcity of Fourier-type gates comes from the non-orientable seam. Locality-preserving gates are constrained by anyon data and compatibility with mapping-class-group actions [15]. The Clifford restriction of Bravyi and Koenig applies to their two-dimensional topological qubit stabilizer-code setting [16]; it is not used here as a theorem for composite NN. Proposition 3 gives the dressed-isometry criteria within its stated class and grid.

Twisted homology of surfaces with cross-caps

Let Σg,k\Sigma_{g,k} (k≥1k\ge1; Σ0,1=RP2\Sigma_{0,1}=\mathbb{R}P^2) be the connected sum of gg tori and kk cross-caps, π1=⟨ai,bi,cj∣r⟩\pi_1=\langle a_i,b_i,c_j\mid r\rangle, r=∏i[ai,bi]∏jcj2r=\prod_i[a_i,b_i]\prod_jc_j^2, coefficients ZN2\mathbb{Z}_N^2, ρ(ai)=ρ(bi)=1\rho(a_i)=\rho(b_i)=1, ρ(cj)=Pηj\rho(c_j)=P^{\eta_j} with P∈{Pd,Pe}P\in\{P_d,P_e\} and the uniform pattern ηj=1\eta_j=1 or the alternating pattern ηj=(−1)j−k\eta_j=(-1)^{j-k} (k even). The Klein bottle of § 5.1 is alternating with k=2k=2 (c1=αβ↦Pc_1=\alpha\beta\mapsto P, c2=β−1↦P−1c_2=\beta^{-1}\mapsto P^{-1}). Let Qk=∑i<kP2iQ_k=\sum_{i<k}P^{2i}.

Theorem 5 (twisted homology). Let g≥0g\ge0, k≥1k\ge1 and 2g+k≥22g+k\ge2.1 Parts (b)–(e) assume the consistency condition in (a).

(a) The local system is consistent iff P2∑jηj=1P^2\sum_j\eta_j=1: for the uniform pattern iff P2k=1P^{2k}=1 (exchange: kk even or N=2N=2; diagonal: u2k=1u^{2k}=1); the alternating pattern always is.

(b) For the uniform pattern,

H1≅(ZN)2g⊕ker⁡[d1:(ZN2)k→ZN2]im⁡[d2:ZN2→(ZN2)k](9)H_1\cong(\mathbb{Z}_N)^{2g}\oplus\frac{\ker\left[d_1:(\mathbb{Z}_N^2)^k\to\mathbb{Z}_N^2\right]}{\operatorname{im}\left[d_2:\mathbb{Z}_N^2\to(\mathbb{Z}_N^2)^k\right]} \tag*{(9)}

d1(m1,…,mk)=∑j(P−1)mjd_1(m_1,\ldots,m_k)=\sum_j(P-1)m_j, d2(m)=((1+P)m,P2(1+P)m,…,P2(k−1)(1+P)m)d_2(m)=((1+P)m,P^2(1+P)m,\ldots,P^{2(k-1)}(1+P)m); for the alternating pattern the jj-th blocks carry PηjP^{\eta_j} and prefactors P2(η1+⋯+ηj−1)P^{2(\eta_1+\cdots+\eta_{j-1})}.

(c) For both patterns

∣H1∣=N2g+k−1∣ker⁡(1+P)∣∣im⁡(P−1)∣(10)|H_1|=N^{2g+k-1}\frac{|\ker(1+P)|}{|\operatorname{im}(P-1)|} \tag*{(10)}

that is, N2(g+k−2)gcd⁡(2,N)2N^{2(g+k-2)}\gcd(2,N)^2 in the exchange branch and N2(g+k−2)(γ−γ+)2N^{2(g+k-2)}(\gamma_-\gamma_+)^2 in the diagonal branch.

(d) H1≅ZN2(g+k−2)⊕coker⁡(1+P)⊕ker⁡(1−P)H_1\cong\mathbb{Z}_N^{2(g+k-2)}\oplus\operatorname{coker}(1+P)\oplus\ker(1-P) for the alternating pattern, and for the uniform one whenever Qk(1+P)≡0(modN)Q_k(1+P)\equiv0\pmod N or NN is prime; in the exchange branch with even kk, Qk=0Q_k=0. These conditions are sufficient, not necessary.

(e) H1(Σg,k)≅ZN4g⊕H1(Σ0,k)H_1(\Sigma_{g,k})\cong\mathbb{Z}_N^{4g}\oplus H_1(\Sigma_{0,k}). For k=1k=1, the auxiliary term on the right is given by Example 1.

Proof. For the one-vertex complex, ∂2m=(ρ(∂r/∂x)m)x\partial_2m=(\rho(\partial r/\partial x)m)_x and ∂1(mx)=∑x(ρ(x)−1)mx\partial_1(m_x)=\sum_x(\rho(x)-1)m_x. The Fox derivatives of the commutators vanish under ρ\rho, and ∂r/∂cj↦P2(η1+⋯+ηj−1)(1+Pηj)\partial r/\partial c_j\mapsto P^{2(\eta_1+\cdots+\eta_{j-1})}(1+P^{\eta_j}), giving (b), (e) and ρ(r)=P2∑ηj\rho(r)=P^2\sum\eta_j, hence (a). As PP is invertible, ker⁡d2=ker⁡(1+P)\ker d_2=\ker(1+P) and im⁡d1=im⁡(P−1)\operatorname{im}d_1=\operatorname{im}(P-1), giving Eq. (10); ∣ker⁡(1±P)∣\lvert\ker(1\pm P)\rvert for Pd,PeP_d,P_e is computed without dividing by non-units. For (d), coordinates (m1,…,mk−1,mˉ)(m_1,\ldots,m_{k-1},\bar m) with mˉ=∑jΛjmj\bar m=\sum_j\Lambda_jm_j (Λj∈{1,−P−1}\Lambda_j\in\{1,-P^{-1}\}) put the mˉ\bar m-component of im⁡d2\operatorname{im}d_2 in the form Qη(1+P)mQ_\eta(1+P)m, with Qη=0Q_\eta=0 (alternating) or QkQ_k (uniform); when it vanishes the quotient splits, and for prime NN the order fixes the structure.

Example 1 (projective plane). [derivation; not on the preregistered grid] For Σ0,1=RP2\Sigma_{0,1}=\mathbb{RP}^2, set M=ZNM=\mathbb{Z}_N and ρ(c)=P\rho(c)=P. The relation c2=1c^2=1 requires P2=1P^2=1. In the same cellular convention as above the complex is

M→1+PM→P−1M,M \xrightarrow{1+P} M \xrightarrow{P-1} M,

so direct calculation from this complex gives

H1(RP2;Mρ)=ker⁡(P−1)im⁡(1+P),H_1(\mathbb{RP}^2;M_\rho)=\frac{\ker(P-1)}{\operatorname{im}(1+P)},
∣H1∣=∣ker⁡(P−1)∣∣ker⁡(1+P)∣N2.\lvert H_1\rvert=\frac{\lvert\ker(P-1)\rvert\lvert\ker(1+P)\rvert}{N^2}.

In the diagonal branch consistency is u2≡1(modN)u^2\equiv1\pmod N, including u=±1u=\pm1 for every NN. With dρ=γ−γ+/Nd_\rho=\gamma_-\gamma_+/N, a positive integer for these consistent cases,

H1≅Zdρ⊕Zdρ,∣H1∣=dρ2.H_1\cong\mathbb{Z}_{d_\rho}\oplus\mathbb{Z}_{d_\rho},\qquad\lvert H_1\rvert=d_\rho^2.

In particular, u=±1u=\pm1 gives dρ=gcd⁡(2,N)d_\rho=\gcd(2,N). In the exchange branch P2=−1P^2=-1, so consistency requires N=2N=2; then H1=0H_1=0 and ∣H1∣=1\lvert H_1\rvert=1. These are homology statements, not a construction of a lattice code or new sealed grid data.

Proof. The one-cell presentation has attaching word c2c^2, whose Fox derivative is 1+c1+c; hence the two differentials above. Their product is P2−1P^2-1, so consistency puts im⁡(1+P)\operatorname{im}(1+P) inside ker⁡(P−1)\ker(P-1). The order follows by taking the quotient and using ∣im⁡(1+P)∣=N2/∣ker⁡(1+P)∣\lvert\operatorname{im}(1+P)\rvert=N^2/\lvert\ker(1+P)\rvert. For PdP_d with u2=1u^2=1, one has u−1=uu^{-1}=u. In its two cyclic components the kernels of P−1P-1 have orders γ+\gamma_+ and γ−\gamma_-, and the images of 1+P1+P have orders N/γ−N/\gamma_- and N/γ+N/\gamma_+, respectively. Each quotient is cyclic of order dρd_\rho; this also proves its integrality. For u=±1u=\pm1, γ−γ+=Ngcd⁡(2,N)\gamma_-\gamma_+=N\gcd(2,N). For PeP_e at N=2N=2, PP swaps the two components and both ker⁡(P−1)\ker(P-1) and im⁡(1+P)\operatorname{im}(1+P) are the diagonal subgroup, so their quotient is trivial.

Scope. Lattice codes exist here only for the alternating local system on the Klein bottle, (g,k)=(0,2)(g,k)=(0,2) (§ 5.1). For every other surface and local system—handles, k=1k=1 or k≥3k\ge3, and the uniform pattern on the Klein bottle itself, whose holonomy along the rows would be P2P^2—

NNν\nubranchpattern(g,k)(g,k)H1H_1dim
21exch.uniform(0,4)(0,4)Z26\mathbb{Z}_2^68
31exch.altern.(0,2)(0,2)001
31exch.uniform(0,3)(0,3)inconsistent—
31exch.uniform(0,4)(0,4)Z34 (n4=9)\mathbb{Z}_3^4\ (n_4=9)9
31exch.uniform(0,6)(0,6)Z39 (n6=81)\mathbb{Z}_3^9\ (n_6=81)81
31exch.altern.(1,2)(1,2)Z34 (n2=1)\mathbb{Z}_3^4\ (n_2=1)9
41exch.uniform(0,4)(0,4)Z22⊕Z4\mathbb{Z}_2^2\oplus\mathbb{Z}_432
52diag.uniform(0,3)(0,3)inconsistent—
72diag.uniform(0,3)(0,3)Z7\mathbb{Z}_77
83diag.uniform(0,3)(0,3)Z8∗\mathbb{Z}_8^*64

Table 4. Twisted homology of Σg,k\Sigma_{g,k} (E-4). dim⁡=∣H1∣1/2\dim=\lvert H_1\rvert^{1/2}, the dimension a code would have if its logical module were H1H_1; a lattice code exists only for the alternating (0,2)(0,2) row (the Klein bottle of § 5.1). nkn_k as in Remark 1. *Reading: not direct-sum form.

Theorem 5 and the separate projective-plane calculation in Example 1 count homology only. A local cross-cap can be realized as a slit whose two lips are glued by a reflection; its ends merge into a vertex of cone angle 4π4\pi, where the compatible gluings would have to be classified anew. If such a code exists and its logical module is the twisted homology, its dimension is ∣H1∣1/2\lvert H_1\rvert^{1/2}, e.g. 9 (two qutrits) for N=3N=3, k=4k=4 in the exchange branch (Table 4); we do not claim that every lattice realization would have this logical module.

Remark 1 (charge-conjugation defects: consistency only). For odd NN, CC fixes only the vacuum, so the counting of symmetry defects [17] gives one CC-defect type DD, with D⊗D=∑(A,B)∈ZN2(A,B)D\otimes D=\sum_{(A,B)\in\mathbb{Z}_N^2}(A,B) and quantum dimension NN (for defects of other anyon symmetries of ZN\mathbb{Z}_N models see Refs. [10, 18]); kk such defects (kk even) have nk=Nk−2n_k=N^{k-2} vacuum channels, and in the exchange branch with even kk, ∣H1∣=(N2gnk)2\lvert H_1\rvert=(N^{2g}n_k)^2 by Eq. (10) (432 grid cases). This is one count written twice: consistent with, not a proof of, cross-caps acting as charge-conjugation defects. For even NN, CC fixes the four labels {0,N/2}2\{0,N/2\}^2, so the same counting gives four CC-defect types, related by fusion with abelian anyons and hence of equal dimension, with ∑D2=N2\sum\mathcal{D}^2=N^2, i.e. d=N/2d=N/2 (derivation, not computed).

Around a cross-cap: the double-winding monodromy

A ZZ-type path from column x0x_0 of row 0 runs through (x0+y,y)(x_0+y,y), y<Lyy<L_y (a ZZ on (x,y)(x,y) fails to commute only with the plaquettes (x,y)(x,y) and (x−1,y−1)(x-1,y-1), of colour x+yx+y), enters the seam plane at what appears its last site and continues, as in the proof of Theorem 2(ii), from its NW far corner at column x0′=a−x0−Ly+1x'_0=a-x_0-L_y+1; a second pass returns it to x0x_0 (Fig. 3a): twice through the seam, once around a cross-cap. XX-type paths run along (−1,1)(-1,1). The closure count is the number of strings on the 2Ly2L_y path sites, with only ZZ (or only XX) powers, that commute with all stabilizers; it equals ∣ker⁡(h−1)∣|\ker(h-1)| for the multiplier hh picked up once around.

Figure 3. (a) ZZ-type double-winding path on 6×36\times3, a=0a=0 (diagonal branch), sites in order. The first pass (filled) runs from x0=0x_0=0; from its last site the string enters the seam plate through the NW far corner, i.e. row 0 at column a−x=4a-x=4 (top labels), where the second pass (open) starts; it re-enters at column 0 =x0=x_0. Here h=ux0′/ux0=u2h=u_{x'_0}/u_{x_0}=u^2. (b) Closure counts of all 1800 registered paths (sealed E-4 results) by NN and branch, diagonal-branch paths split by whether u2≡1u^2\equiv1: each group has one value, gcd⁡(u2−1,N)\gcd(u^2-1,N) (diagonal) or gcd⁡(2,N)\gcd(2,N) (exchange). Grey line: NN.

Proposition 4 (double-winding monodromy). For all 540 codes with N≤12N\le12, all units, (Lx,Ly)∈{(4,3),(4,4),(6,3)}(L_x,L_y)\in\{(4,3),(4,4),(6,3)\}, a≤3a\le3, and all 1800 registered paths, the closure count is gcd⁡(u2−1,N)\gcd(u^2-1,N) in the diagonal and gcd⁡(2,N)\gcd(2,N) in the exchange branch. The multiplier read generator by generator is h∈{u2,u−2}h\in\{u^2,u^{-2}\} on all 720 diagonal-branch paths and h=−1h=-1 on all 1080 exchange-branch paths; for N>2N>2, h=−1h=-1 occurs in the diagonal branch exactly for u2≡−1u^2\equiv-1 (64 paths: N=5N=5, u=2,3u=2,3; N=10N=10, u=3,7u=3,7).

The closure counts were pre-registered; hh is a reading, equal on all paths to h=ux0′ux0h=u_{x'_0}u_{x_0}: the landing columns sum to a−Ly+1a-L_y+1 and have equal parity in the diagonal branch (h=u±2h=u^{\pm2}) and opposite parity in the exchange branch (h=u⋅(−1/u)=−1h=u\cdot(-1/u)=-1).

These readings concern the external monodromy of the defined strings: on the registered lattices (derivation and reading), a charge carried once around a cross-cap is acted on by P±2P^{\pm2}, the sign depending on the direction—CC in the exchange branch, the multiplier u±2u^{\pm2} in the diagonal branch (CC if u2≡−1u^2\equiv-1). They do not determine the defect type, the cross-cap states, or FF and RR data (Remark 1).

Cross-caps and magic: a conditional boundary

A stabilizer state of nn qudits is, up to phase, the unique common eigenvector of an abelian group of NnN^n Weyl operators, i.e. of a Lagrangian subgroup G⊂ZN2nG\subset\mathbb{Z}_N^{2n} (G⊥=GG^\perp=G) with phases; a logical stabilizer state is a code state fixed by a maximal abelian group of logical Weyl operators.

Corollary 1 (conditional no-go). In the operation model O\mathcal{O} every operation maps stabilizer states to stabilizer states, and the outcome statistics of any protocol in O\mathcal{O} applied to a stabilizer input can be computed efficiently classically. Hence, within O\mathcal{O}, cross-cap geometry—whatever the seam action PP, the number of cross-caps and the

monodromy—produces no non-stabilizer state from stabilizer inputs, and no protocol in O\mathcal{O} maps a logical stabilizer state of a code of this type to a non-stabilizer state.

Proof. This is the qudit Gottesman–Knill theorem [19, 20, 21, 22] restricted to O\mathcal{O}; we check the steps, including composite NN. Clifford unitaries, among them qudit permutations and single-site dressings, map Weyl operators to Weyl operators up to phases, hence Lagrangian groups to Lagrangian groups. Measuring W(w)W(w) with outcome ξ\xi on the state of G\mathcal{G} gives, if ξ\xi has non-zero probability, a state stabilized by the elements of G\mathcal{G} that commute with W(w)W(w) and by ξ−1W(w)\xi^{-1}W(w), i.e. by G′=(G∩w⊥)+⟨w⟩\mathcal{G}'=(\mathcal{G}\cap w^\perp)+\langle w\rangle. G′\mathcal{G}' is isotropic, and since (A∩B)⊥=A⊥+B⊥(A\cap B)^\perp=A^\perp+B^\perp for subgroups of ZN2\mathbb{Z}_N^2 and ⟨w⟩⊂w⊥\langle w\rangle\subset w^\perp, the modular law gives G′⊥=(G+⟨w⟩)∩w⊥=G′\mathcal{G}'^\perp=(\mathcal{G}+\langle w\rangle)\cap w^\perp=\mathcal{G}' is Lagrangian. Feed-forward corrections are Weyl or Clifford operations, and tracking G\mathcal{G} with its phases is efficient [22]. A logical stabilizer state is a stabilizer state: the preimage in S⊥\mathcal{S}^\perp of a Lagrangian subgroup of L\mathcal{L} is isotropic of order ∣S∣∣L∣1/2=Nn|\mathcal{S}||\mathcal{L}|^{1/2}=N^n.

Remark 2 (Boundary of the operation model). Corollary 1 applies only to protocols in O\mathcal{O}. Coherent path superpositions of mobile charge-conjugation defects and their interferometric measurements, as used in a different construction [9], are outside the present scope and are not implemented here. Gauging a separately specified charge-conjugation symmetry, when consistent symmetry and defect data permit it [17], changes the theory and is outside the present scope. Reflection-enriched data, fractionalization and cross-cap states [5] are outside the present scope and are not fixed by seam monodromy alone. In particular, the additional symmetry structure supporting the charge measurement considered by Cong, Cheng and Wang [8] is not constructed from our gluing. External magic states [23] are excluded by assumption: protocols supplied with them are outside the present claim. Moving a defect by the Clifford code deformations already allowed in O\mathcal{O} does not by itself supply coherent path control. The label relation Pe2=CP_e^2=C alone supplies none of these additional structures. These exclusions neither establish impossibility outside O\mathcal{O} nor demonstrate any such resource in the lattice codes studied here.

The closure argument is the standard stabilizer argument; the contribution here is its geometric setting and boundary. Theorem 1 classifies commuting gluings within the local monomial far-corner ansatz, and Proposition 3 identifies the dressed isometries on the registered grid within its single-site dressing class. For every realization and protocol satisfying O\mathcal{O}, the obstruction is uniform in the seam action, the number of cross-caps and the monodromy, because these parameters do not alter stabilizer closure under the permitted operations. This does not assert that arbitrary physical motions or measurements of cross-caps admit an implementation in O\mathcal{O}. The argument needs no classification of locality-preserving gates [16] (we neither use nor extend that result to composite NN or non-orientable codes).

What the post-measurement states show

For odd NN we use the Wigner function of Gross [24], defined by Wρ(q,p)=Tr⁡[ρA(q,p)]/NW_{\rho}(q,p)=\operatorname{Tr}[\rho A(q,p)]/N and A(q,p)∣j⟩=ω2p(j−q)∣2q−j⟩A(q,p)|j\rangle=\omega^{2p(j-q)}|2q-j\rangle, so that A(0,0)=CA(0,0)=C. Let ψ±(j)=(∣j⟩±∣−j⟩)/2\psi_{\pm}(j)=(|j\rangle\pm|-j\rangle)/\sqrt{2} (j≠0j\ne0) and Nj=N/gcd⁡(j,N)N_j=N/\gcd(j,N); for prime NN and λ=ω2≠1\lambda=\omega^2\ne1 of order rr, let ϕ0,ℓ=r−1/2∑m<re−2πiℓm/r∣my⟩\phi_0,\ell=r^{-1/2}\sum_{m<r}e^{-2\pi i\ell m/r}|m y\rangle be the eigenstates of μλ∣x⟩=∣λx⟩\mu_\lambda|x\rangle=|\lambda x\rangle on an orbit O={λmy}⊂ZN×O=\{\lambda^m y\}\subset\mathbb{Z}_N^\times.

Proposition 5 (Wigner negativity). (a) The minimum of WW is −1/N-1/N for ψ−(j)\psi_-(j), at (0,0)(0,0), and −cos⁡(π/Nj)/N-\cos(\pi/N_j)/N for ψ+(j)\psi_+(j). (b) Measuring CC on ∣j⟩|j\rangle gives ψ±(j)\psi_\pm(j) with probability 1/21/2 each; measuring Z+Z†Z+Z^\dagger on ∣+⟩|+\rangle gives ∣0⟩|0\rangle with probability 1/N1/N and ψ+(j)\psi_+(j) with probability 2/N2/N, j=1,…,(N−1)/2j=1,\ldots,(N-1)/2. (c) For prime NN, min⁡Wϕℓ≤−1/((N−1)r2)\min W_{\phi_\ell}\le-1/((N-1)r^2), and measuring μλ\mu_\lambda on ∣x⟩|x\rangle, x∈Ox\in O, gives each eigenvalue with probability 1/r1/r.

Proof. (a), (b) by direct computation: WW vanishes off q∈{0,±j}q\in\{0,\pm j\}, W(0,p)=±cos⁡(4πpj/N)/NW(0,p)=\pm\cos(4\pi pj/N)/N, W(±j,p)=1/(2N)W(\pm j,p)=1/(2N). (c) A subset of ZN\mathbb{Z}_N (NN prime) with at least two elements that is closed under midpoints is all of ZN\mathbb{Z}_N; as 0∉O0\notin O, some x≠x′∈Ox\ne x'\in O have their midpoint q∉Oq\notin O. In that column ∑pW(q,p)=ρqq=0\sum_p W(q,p)=\rho_{qq}=0, ∑pW(q,p)2=mq/(Nr2)\sum_p W(q,p)^2=m_q/(Nr^2) with mq=∣O∩(2q−O)∣≥2m_q=|O\cap(2q-O)|\ge2, and ∣W∣≤1/N|W|\le1/N, so the at most N−1N-1 negative entries give min⁡W≤−mq/(2(N−1)r2)\min W\le-m_q/(2(N-1)r^2).

A pure state of odd dimension has a non-negative Wigner function iff it is a stabilizer state [24], and negativity is necessary for magic-state distillation to help [25]; ψ−(1)\psi_-(1) at N=3N=3 is the strange state, min⁡W=−1/3\min W=-1/3 [23]. Proposition 5 therefore certifies only that one non-stabilizer measurement—of CC, or of a charge that does not distinguish a label from its conjugate, such as Z+Z†Z+Z^\dagger—yields non-stabilizer states with fixed probabilities. That measurement lies outside O\mathcal{O}. Nothing here shows distillability, fault tolerance, or a universal gate set.

Symmetry protection, by hand

Could the orientation reversal itself protect a measurement that does not distinguish a charge from its conjugate? Take a tunnelling term Ht=−(tV+t∗V†)H_t=-(tV+t^*V^\dagger). The invariance of the Hamiltonian can be maintained in three ways. (i) A unitary CC with CVC−1=V†CVC^{-1}=V^\dagger requires t=t∗t=t^*. (ii) An antiunitary TT also conjugates scalars: if TVT−1=V†TVT^{-1}=V^\dagger, HtH_t is invariant for every complex tt; only if VV is TT-even, TVT−1=eiϑVTVT^{-1}=e^{i\vartheta}V, does invariance give t=eiϑt∗t=e^{i\vartheta}t^*. This is the mechanism of Cong, Cheng and Wang: in their fractional quantum spin Hall realization “the time-reversal symmetry exchanges the two layers”, the charge ee is time-reversal invariant, and so “the tunneling amplitude of ee has to be real”; “the requirement that tt is real is beyond topological protection” [8]. (iii) A unitary spatial reflection that maps the coupling path to itself and VV to V†V^\dagger pins tt exactly as CC does: on the literal gluing UσFpUσ†=Fσ(p)U_{\sigma}F_pU_{\sigma}^{\dagger}=F_{\sigma(p)} (Proposition 1), so a σ\sigma-invariant term −(tFp+t∗Fp∗)-(tF_p+t^*F_p^*) on a σ\sigma-fixed plaquette (y=(Ly−1)/2(y=(L_y-1)/2 and 2x≡a−1(modLx))2x\equiv a-1\pmod{L_x}), e.g. x=0,2x=0,2 at 4×34\times3, a=1)a=1) has tt real.

The Klein-bottle gluing, by contrast, is a gluing rule, not a symmetry of the Hamiltonian, and by itself constrains no local amplitude. The three pinning symmetries are microscopic symmetries that must be imposed; they are equally available on orientable lattices (a mirror through the coupling on a plane or torus) and fragile to disorder. If a Klein-bottle lattice carries such a reflection, its protection is symmetry protection, not topological, and does not come from non-orientability. Within O\mathcal{O} the answer is negative; Corollary 1 and O\mathcal{O} are unchanged. This is a hand assessment; the exclusions of O\mathcal{O} apply, among them reflection-enriched data [5] and path superpositions of mobile defects [9].

Relation to neighbouring work

Homological codes. Lei shows that a qudit homological code on any 2-complex, for any qudit dimension DD, has dimension ∣H1(Σ;ZD)∣|H_1(\Sigma;\mathbb{Z}_D)| [3]; our edge-register codes are instances. Chan, Teo and Ryu relate the degeneracy on non-orientable surfaces to parity-invariant anyons [4]; Theorem 2 establishes this count, ∣ker⁡(1−P)∣|\ker(1-P)|, for the seam actions of local lattice gluings, including the order-four ones. Bombin and Martin-Delgado state that “for nonorientable surfaces it is impossible to construct homological codes based on qudits of dimension D>2D>2” [26], tracing this to “the torsion subgroup” Z2\mathbb{Z}_2 of H1H_1 through their requirement that the stabilizer subspace have “a basis which is a linearly independent set” in ZD2n\mathbb{Z}_D^{2n}, so that the code is an integer number of DD-dimensional qudits (their example: D=4D=4 on the projective plane). For odd DD that torsion is absent (H1(K;ZD)=ZDH_1(K;\mathbb{Z}_D)=\mathbb{Z}_D). Their no-go concerns their construction, a free linear subspace of ZD2n\mathbb{Z}_D^{2n}. Codes whose stabilizer is a module, which allows ZN\mathbb{Z}_N torsion, exist on non-orientable surfaces for every DD: Lei’s theorem gives them in general, and explicitly the edge register on KK, the toric code of a cellulation, has commuting stabilizers and, for odd NN, dimension NN. What fails is therefore the broad reading that no non-orientable code with D>2D>2 exists; we do not claim that the two constructions coincide. For even DD the obstruction concerns only the integer-number-of-qudits requirement, which the Zgcd⁡(2,D)\mathbb{Z}_{\gcd(2,D)} torsion violates; our edge-register codes (dimension 2N2N) do not impose it and are the twisted homology of Proposition 2 with P=diag⁡(1,−1)P=\operatorname{diag}(1,-1). The site-register uu-codes are not cellular codes in the usual sense (their checks change type and carry unit multipliers across the seam), but on the grid their logical modules are twisted homology groups as well.

Cross-cap gates and reflection-enriched order. Cross-cap defects implement a fault-tolerant logical Hadamard in the Z2\mathbb{Z}_2 surface code [6]; Proposition 3 is a ZN\mathbb{Z}_N comparison for lattice isometries. Barkeshli, Bonderson, Cheng, Jian and Walker develop reflection- and time-reversal-enriched phases on non-orientable manifolds, with cross-cap states and fractionalization data [5]. Their Sec. II acknowledges order-four and higher reflection actions before adopting an order-two convention for the subsequent analysis. Our exchange-branch P2=CP^2=C is an explicit local-system monodromy of the stated lattice gluing; we do not claim order-four reflection itself as new or identify our cross-caps with their defect classes.

Quotient codes and twisted compactification. Hsin [7] formulates monomial stabilizer-code quotients through twisted compactification, with a transfer theorem for logical operators and distance, a Clifford restriction for descended monomial symmetries, and a separate discussion of composite dimensions. Our focus is the explicit local far-corner seam ansatz of the ZN\mathbb{Z}_N plaquette model: its commuting uu-family, phase-sensitive consistency, and the noncommuting literal seam with bulk-locked clock spectra. These are model-resolved questions, rather than a general quotient transfer theorem. The non-Clifford constructions of Ref. [7] use additional cup-product circuits or fermionic pumps outside O\mathcal{O}; they do not contradict the stabilizer-closure statement of Corollary 1.

Clifford limits and routes to magic. Webster and Bartlett prove non-universality for combinations of defect braiding and locality-preserving logical operators, including Abelian quantum-double models; their stronger Clifford restriction applies to a specified standard defect encoding [27]. Corollary 1 instead follows directly from the explicitly chosen stabilizer operation model. In the charge-4e superconductor of Shi, Han, Raghu and Vishwanath the hc/4ehc/4e fluxes are charge-conjugation defects of a chiral Z3\mathbb{Z}_3 order: braiding gives the many-qutrit Clifford group, and a single-probe interferometric measurement prepares magic states [9]. Our exchange-branch cross-caps also have double-winding monodromy CC, but within O\mathcal{O} a cross-cap moves only by code deformation, and no coherent superposition of paths arises.

Other neighbours. Lattice twists on a Klein bottle must represent its fundamental group [28]. Takahashi, Li and Yoshida construct an order-two locality-preserving Clifford realization of electric–magnetic duality for odd NN and exclude it for even NN on periodic toric-code lattices sufficiently large compared with the locality range; their obstruction locally emulates a cross-cap [29]; Proposition 3 asks instead which isometries of the Klein-bottle codes admit single-site dressings; its condition for a non-trivial colour exchange, u2≡−1u^2 \equiv-1, comes from compatibility with the seam action. Self-dual chains appear in parafermion chains [13], twist-defect chains [30] and edges of Wen’s model [31]; lattice defects implement Kramers–Wannier duality in the Ising model [32], and lattice reflections implement it in one dimension [33]; specific here is its realization by the Klein-bottle reflection σ\sigma exchanging the two plaquette rows of an obstructed gluing.

Discussion

The classification covers local, monomial, far-corner gluings only; non-local gluings, multi-site far corners, modified bulk plaquettes and the glide gluings of Ref. [1] at N>2N > 2 were not studied. Grid statements are not extrapolated (the odd-NN edge distance dX=Lxd_X = L_x fails beyond the grid); the site-register logical module is verified on the grid and only argued in general; the gate results hold within the single-site dressing class; lattice codes exist only for the alternating Klein bottle, and elsewhere we only count homology; the ring degeneracy is not claimed to be protected. The no-go holds within O\mathcal{O}, and the symmetry assessment is by hand. Producing a non-stabilizer resource from stabilizer inputs requires an operation outside O\mathcal{O}; the extensions listed in Remark 2 are outside the present scope, and their availability or sufficiency in these lattice codes is not established here. Thresholds, decoders and noise were not considered.

Acknowledgments

The research program was designed and directed by the author. Derivations, pre-registered predictions, code, figures and drafting were produced with AI assistance (Anthropic Claude). Independent AI instances, including OpenAI Codex, assisted with cross-checks; these were AI-assisted checks, not independent human verification or journal peer review. The reports and dispositions are archived with the accompanying material (Appendix A). The author takes full responsibility for the content. This work received no external funding.

Evidence, provenance and reproducibility

Registered grids

E-1: N∈{2,…,13,16,17,25,26}N \in\{2,\ldots,13,16,17,25,26\}, Lx∈{3,…,8}L_x \in\{3,\ldots,8\}, Ly∈{3,4,5}L_y \in\{3,4,5\}, a≤3a \le3, all units. E-2: N∈{2,…,10,12}N \in\{2,\ldots,10,12\}, Lx∈{3,…,8}L_x \in\{3,\ldots,8\}, Ly∈{2,…,5}L_y \in\{2,\ldots,5\}, a<min⁡(Lx,4)a < \min(L_x,4); 920 cases. E-3: site register N∈{2,…,12,15,16}N \in\{2,\ldots,12,15,16\}, all uu, Lx∈{4,6}L_x \in\{4,6\}, Ly∈{3,4,5}L_y \in\{3,4,5\}, a≤3a \le3, plus 8×38 \times3, a≤1a \le1 (1586 codes); edge register N≤12N \le12, Lx∈{3,…,6}L_x \in\{3,\ldots,6\}, Ly∈{3,4,5}L_y \in\{3,4,5\}, a<min⁡(Lx,4)a < \min(L_x,4) (495 codes); isometries on 1378 site (N≤12N \le12 and 16) and 264 edge codes (Lx∈{4,6}L_x \in\{4,6\}); distances on 135 site and 30 edge codes. E-4: Sec. 5.3, Propositions 4, 5.

Pre-registration and runs

E-0 is a calibration without pre-registration. For E-1 to E-4 the predictions, anchors, grids and falsification criteria were frozen in prereg/eK_prereg.md before the stage’s runs in this package (sha256 and time in logs/eK_prereg_stamp.txt, Table 5; records, not cryptographic proofs of priority) and were not changed afterwards. Each lists what was known beforehand, which is not counted as prediction: for E-1, E-0 readings on the Fourier-dressed code (code dimensions at 4×34 \times3, 4×44 \times4, N≤6N \le6, part of H3; the N=2N = 2 odd-LxL_x sign, part of H4); for E-2, the literal-gluing anatomy and the N=2N = 2 results; for E-3, the E-1 results and the N=2N = 2 codes; for E-4, the seam action PP, planning-stage checks by another AI instance outside the package (∣H1∣=N2(k−2)|H_1| = N^{2(k-2)} for kk identical exchange-branch cross-caps on the sphere, N=3,5,7,9,15N = 3,5,7,9,15, even k≤6k \le6; the N=3N = 3 Wigner minima −1/6-1/6, −1/3-1/3) and a hand computation of the uniform N=8N = 8, u=3u = 3, k=2k = 2 case, which overlap H1(b) and H4(a). The seam action PP, P2=CP^2 = C and the phase statement of Theorem 2(iii) were not pre-registered (E-1 registered ∣R∣|\mathcal{R}| and the phase-free frustration rule); PP first appears in a pre-registration as the stated basis of E-3 H1.

stageprereg frozensealsealed
E-0(calibration)0901726bc670b9df10-01 10:11:53
E-109-30 15:20:3755136031bc669cbb10-01 09:52:44
E-210-01 14:49:29599bdc643a1cdd7d10-02 07:41:10
E-310-02 09:04:58cf526247f493155310-02 11:10:21
E-410-02 17:45:4497811bc292074cfe10-02 18:35:22
checks:E-1 21 108/10 027/2 707; E-2 15 956/13/1 410;
E-3 25 656/2 408/6 698; E-4 8 798/2 121/7 330

Table 5. Pre-registration freeze, seal (first 16 hex digits; full values in SEALS.txt) and sealing time of each stage (UTC; all dates in 2026); checks are prediction/anchor/cross-check items of the sealed run.

Each sealed run is one execution of code/run_eK.sh that started with no result files of its stage (the manifest lists the files present at the start). It was preceded by full development or rehearsal runs, whose results were inspected and in part reviewed, and code, gates and check counts changed between them. For example, E-0 ran twice against the v0.4 targets of Ref. [1] on 2026-09-30 and again on 2026-10-01; E-1 ran on 2026-09-30 at 16:37 UTC (57 018 checks; reviewed) and 19:21 (59 042); E-2 was sealed, reviewed and rerun with a changed gate and identical checks (superseded seal 6d0c3e1f…); E-3 had two rehearsals (the reviewed one with 36 329 checks; sealed run 34 762); after E-4’s reviewed rehearsal, 1800 cross-checks of hh were added.

Evidence by result

The proofs of draft v0.1 were re-derived by an independent AI reviewer; those added in v0.2 (Corollary 1, the edge-register part of Proposition 2) were checked by the external AI review of draft v0.2 but have not been re-derived independently by a human. Sealed checks per hypothesis (correlated within a case): E-1 H1–H4 4182, 261, 4536, 9252; E-2 H1–H5 3105, 7600, 196, 399, 6040; E-3 H1–H5 7531, 11 566, 390, 4926, 4602; E-4 H1–H4 5720, 864, 5400, 414. Theorem 1: complete enumeration of all far-corner assignments in 16 small cases (644 782 854 gluings: N=2N = 2, Lx≤6L_x \le6; N=3N = 3, Lx≤4L_x \le4; N=4N = 4, 5, Lx=3L_x = 3, Ly=3L_y = 3, a≤1a \le1) finds exactly the uu-family; a pruned exact search returns the predicted sets on all 1152 grid cases; for N≤30N \le30 (Lx=3,5L_x = 3, 5) odd-LxL_x gluings exist exactly for N∈{2,5,10,13,17,25,26,29}N \in\{2, 5, 10, 13, 17, 25, 26, 29\}. Theorem 2: 7272 stabilizer-group orders agree between three algorithms (Howell form [34], integer determinant, Smith form); state-vector ranks equal the algebraic dimension in 21 cases up to 16 777 216 states. Part (ii) rests on 17 208 sealed property checks of the seam map (colour exchange iff a+Lya+L_y even; entries multiplying to −1-1; square CC exactly in the exchange branch and for u2≡−1u^2 \equiv-1; ∣R∣|\mathcal{R}| fixed labels); the equality of the 239 stored matrices with Eq. (3) was checked post hoc; the overall sign, a convention, is not tested. The phase statement of (iii) is a cross-check added after the internal review (846 cases). Theorems 3, 4, Proposition 1: 920 grid cases, including the dimensions of 186 712 joint eigenspaces (54 613 allowed); complete spectra on nine dense lattices (up to 6561 states) and block-resolved at N=3N = 3, 3×33 \times3; Lanczos ground energies and low levels on five lattices (up to 1 953 125 states), ground energies only on three (up to 16 777 216); all 34 ground energies agree with Eq. (5) to 9.5×10−149.5 \times10^{-14}, which also reproduces the six energies and twenty degeneracies of Ref. [1]. The ring identity was checked on 68 ring sizes (N≤8N \le8). Two registered tests were not run (low levels of the three largest lattices; complete ring spectra beyond Nm=2048N^m = 2048), so (c) is confirmed with reduced coverage. Propositions 2, 3: invariant factors of all 1586 site and 495 edge codes by two routes (Smith form of the commutation Gram matrix; Howell counting), with ∣L∣|\mathcal{L}| the

square of the dimension of Theorem 2; gates exhaustive on 1378 site and 264 edge codes. For odd NN, (b) and (c) of Proposition 3 follow from (a), so its informative tests are at even NN; at N=2N = 2 the groups of Ref. [1] are reproduced in its 30 calculators. Theorem 5: 2074 cases, 1570 consistent, invariant factors by two routes; (d) holds on its 1282 cases (550 with NN prime, implied by the order). The 288 consistent uniform cases outside its conditions (all diagonal, composite NN) are ranges: 252 of direct-sum form, 36 not (N=8N = 8: 10; N=16N = 16: 26), e.g. Z4\mathbb{Z}_4 at N=8N = 8, u=3u = 3, (0,3)(0,3), not Z22⊕Z42⊕Z2\mathbb{Z}_2^2 \oplus\mathbb{Z}_4^2 \oplus\mathbb{Z}_2; a post-hoc recomputation from Eq. (9) reproduces all 288. The alternating (0,2)(0,2) case reproduces the modules of all E-3 codes; H2 follows from H1(b). Propositions 4, 5: the closure counts separate the branches on 1120 paths; hh, added after the internal review because closure counts cannot tell h=−1h = -1 from u2u^2 when gcd⁡(2,N)=gcd⁡(u2−1,N)\gcd(2,N) = \gcd(u^2 - 1,N), is a reading; torus control: 28 checks. Wigner: the least negative orbit state is N=13N = 13, Lx=4L_x = 4 (−0.0295-0.0295 against the bound −0.0023-0.0023); stabilizer states have W≥−7.4×10−17\mathcal{W} \ge-7.4 \times10^{-17}. Post hoc (paper/verify_v02.py, exact arithmetic, unsealed): Lemma 1 (N=2N = 2, Lx≤12L_x \le12, Ly≤6L_y \le6, all aa; ∏pFpN/2\prod_p F_p^{N/2} for N≤26N \le26), the dressed glide and the exponents of Theorem 1, Theorem 4(a) in both orders, the CSS form used for Proposition 2, and the σ\sigma-fixed plaquettes of § 6.2.

Engine, seals and gates

A NumPy-only engine represents operators as Weyl strings with exact phases and uses the Howell form [34] for orders, relations and membership, and dense diagonalization and Lanczos for spectra. E-0 calibrated it at N=2N = 2 against the six result files of Ref. [1] (v0.7, unchanged in v0.8): 8005 comparisons (606 stored as truth values), 1934 definitional items, 453 internal cross-checks, 12 self-tests; errors specific to N>2N > 2 are addressed only by each stage’s self-tests and anchors.

Each sealed run writes the sha256 of every code file (gate and seal.py included), input and result to results/eK_files.sha256, whose own sha256 is the seal recorded in SEALS.txt after the gate passed (Table 5); both can be checked with sha256sum alone. The gate code/eK-gate.py re-evaluates every stored comparison, fixes the number and kinds of checks and, for E-2 to E-4, the coverage case by case; the E-3 and E-4 gates recompute 5196 and 13 787 expected values from the preregistered formulas. The gates guard against accidents, not against a writer who rewrites results and seal list consistently; mutation tests (unsealed record) behaved as expected. The gate summaries results/eK_gate.json are not sealed, and running a gate rewrites them. The only complete check is a rerun.

Deviations. Besides the two E-2 tests not run (Appendix § 10.§ 3), the E-2 ring range N≤8N \le8 was fixed in code, not in the pre-registration, and 110 of the 7988 off-grid E-1 cross-checks duplicate a repeated random draw (no verdict affected). Grid-only statements, readings and post-hoc checks are labelled where they occur.

Reviews. Before sealing, an AI instance without the authoring context reviewed each of E-1 to E-4 adversarially, re-implementing the central computations and attacking the gates; its findings (gate defects, two rated P0 in E-1; coverage gaps; an off-grid distance failure; uninformative checks; the blindness of closure counts to hh) were fixed before sealing or recorded. OpenAI Codex reviewed every stage. Its first E-0 review found two P0 problems—the gate trusted cached result files, and the odd-LxL_x dressed gluing of Ref. [1] is sign-frustrated rather than a code with GSD = 2—which led to the seal mechanism, the reruns of E-0 and E-1, and the erratum v0.7 of Ref. [1]; on the current seals it found no P0. The external review of draft v0.2 identified a P1 error in the projective-plane domain of Theorem 5; the v0.3 disposition incorrectly summarized that review as having no P1. Draft v0.4 restores 2g+k≥22g+k \ge2 and separates Example 1, without changing the sealed results. The correction and the preserved v0.3 files are documented in ERRATA v0.4.md; the earlier wording fixes remain in place. For E-2 its point on case-by-case gate coverage led to a changed gate and a full rerun; its other P1 points were adopted. Reports and dispositions are under reviews/, E_campaign/review/ and, for this paper, paper/review_v0.1/ and paper/review_codex_v0.2/.

Data availability. The accompanying ancillary archive, paperE_reproducibility_v0.6.zip, contains the sealed E-0–E-4 code, inputs, results and supporting records in their campaign-relative layout, together with the number-tracing and post-hoc verification scripts. Historical reports retain their original statements; the scientific scope and corrections in this manuscript take precedence. Version DOI for the full reproducibility archive: https://doi.org/10.5281/zenodo.24185575.

Reproduction. bash code/run_eK.sh reruns a stage (sealed wall times 66, 61, 59, 10 min and 7 s for E-0 to E-4 on a two-core machine; Python 3.11.15, NumPy 2.4.4; about 4 GB for E-1, E-2) and gives a new seal; rerunning E-3 invalidates the E-4 seal. paper/claim_numbers.md traces every number, paper/check_numbers.py prints them from the sealed files after verifying the seals (marking unsealed inputs), paper/make_figs.py draws Figure 2–Figure 4, and paper/verify_v0.2.py holds the post-hoc checks. No second human has checked derivations or computations.

Figure 2. Structure of H1(Σg,k;ZN2)H_1(\Sigma_{g,k};\mathbb{Z}_N^2), uniform pattern, diagonal branch, composite NN (columns: units uu grouped by NN; rows: (g,k)(g,k); sealed E-4 results). Blue: Qk(1+P)≡0Q_k(1+P)\equiv0, direct sum confirmed. Green: the condition fails, the direct sum holds (readings). Orange, ×\times: not a direct sum; the extension of Eq. (9) (readings; post-hoc check). Grey: inconsistent, u2k≠1u^{2k}\ne1. The alternating pattern, the exchange branch and prime NN give direct sums throughout the grid. None of these local systems is realized by a lattice code here.

Figure 4. Wigner negativity of post-measurement states (sealed E-4 results; NN odd). (a) −min⁡W-\min\mathcal{W} of ψ±(j)\psi_{\pm}(j) for the 28 registered (N,j)(N,j), against 1/N1/N and cos⁡(π/N)/N\cos(\pi/N)/N; the triangles below the orange line are the four cases with Nj<NN_j<N (two coincide at N=15N=15). (b) Multiplier eigenstates for prime NN (116 states) against the bound 1/((N−1)r2)1/((N-1)r^2): all above the line, by a factor ≥2.59\ge2.59.

References

  1. [1]J. Wang. An exactly solvable obstructed seam on a lattice Klein bottle: bulk-locked Majorana rings and register-dependent toric codes, 2026. URL https://arxiv.org/abs/2610.35931. Version 1.
  2. [2]X.-G. Wen. Quantum orders in an exact soluble model. Phys. Rev. Lett., 90:016803, 2003. doi: 10.1103/PhysRevLett.90.016803.DOI
  3. [3]Z. Lei. Qudit surface codes and hypermap codes. Quantum Inf. Process., 22:297, 2023. doi: 10.1007/s11128-023-04006-8.arxiv.org/abs/2112.01752
  4. [4]A. P. O. Chan, J. C. Y. Teo, and S. Ryu. Topological phases on non-orientable surfaces: twisting by parity symmetry. New J. Phys., 18:035005, 2016. doi: 10.1088/1367-2630/18/3/035005.DOI
  5. [5]M. Barkeshli, P. Bonderson, M. Cheng, C.-M. Jian, and K. Walker. Reflection and time reversal symmetry enriched topological phases of matter: path integrals, non-orientable manifolds, and anomalies. Commun. Math. Phys., 374:1021–1124, 2020. doi: 10.1007/s00220-019-03475-8.DOI
  6. [6]R. Kobayashi and G. Zhu. Cross-cap defects and fault-tolerant logical gates in the surface code and the honey-comb Floquet code. PRX Quantum, 5:020360, 2024. doi: 10.1103/PRXQuantum.5.020360.DOI
  7. [7]P.-S. Hsin. Twisted compactification, quotient quantum codes and logical gates, 2026. URL https://arxiv.org/abs/2611.09538. Version 1.
  8. [8]I. Cong, M. Cheng, and Z. Wang. Universal quantum computation with gapped boundaries. Phys. Rev. Lett., 119:170504, 2017. doi: 10.1103/PhysRevLett.119.170504.DOI
  9. [9]Z. D. Shi, Z. Han, S. Raghu, and A. Vishwanath. Charge-4e superconductors with parafermionic vortices: A path to universal topological quantum computation, 2026. URL https://arxiv.org/abs/2602.09663v2. Version 2.
  10. [10]Y.-Z. You and X.-G. Wen. Projective non-Abelian statistics of dislocation defects in a Z_N rotor model. Phys. Rev. B, 86:161107(R), 2012. doi: 10.1103/PhysRevB.86.161107.DOI
  11. [11]A. Y. Kitaev. Fault-tolerant quantum computation by anyons. Ann. Phys., 303:2, 2003. doi: 10.1016/S0003-4916(02)00018-0.DOI
  12. [12]D. Delmastro and J. Gomis. Symmetries of Abelian Chern–Simons theories and arithmetic. JHEP, 03:006, 2021. doi: 10.1007/JHEP03(2021)006.DOI
  13. [13]W. Li, S. Yang, H.-H. Tu, and M. Cheng. Criticality in translation-invariant parafermion chains. Phys. Rev. B, 91:115133, 2015. doi: 10.1103/PhysRevB.91.115133.DOI
  14. [14]G. Sun, T. Vekua, E. Cobanera, and G. Ortiz. Phase transitions in the Z_p and U(1) clock models. Phys. Rev. B, 100:094428, 2019. doi: 10.1103/PhysRevB.100.094428.DOI
  15. [15]M. E. Beverland, O. Buerschaper, R. Koenig, F. Pastawski, J. Preskill, and S. Sijher. Protected gates for topological quantum field theories. J. Math. Phys., 57:022201, 2016. doi: 10.1063/1.4939783.DOI
  16. [16]S. Bravyi and R. König. Classification of topologically protected gates for local stabilizer codes. Phys. Rev. Lett., 110:170503, 2013. doi: 10.1103/PhysRevLett.110.170503.DOI
  17. [17]M. Barkeshli, P. Bonderson, M. Cheng, and Z. Wang. Symmetry fractionalization, defects, and gauging of topological phases. Phys. Rev. B, 100:115147, 2019. doi: 10.1103/PhysRevB.100.115147.DOI
  18. [18]M. Barkeshli, C.-M. Jian, and X.-L. Qi. Twist defects and projective non-Abelian braiding statistics. Phys. Rev. B, 87:045130, 2013. doi: 10.1103/PhysRevB.87.045130.DOI
  19. [19]D. Gottesman. The Heisenberg representation of quantum computers. arXiv:quant-ph/9807006, 1998.arxiv.org/abs/quant-ph/9807006
  20. [20]D. Gottesman. Fault-tolerant quantum computation with higher-dimensional systems. Chaos Solitons Fractals, 10:1749, 1999. doi: 10.1016/S0960-0779(98)00218-5.DOI
  21. [21]E. Hostens, J. Dehaene, and B. De Moor. Stabilizer states and Clifford operations for systems of arbitrary dimensions and modular arithmetic. Phys. Rev. A, 71:042315, 2005. doi: 10.1103/PhysRevA.71.042315.DOI
  22. [22]N. de Beaudrap. A linearized stabilizer formalism for systems of finite dimension. Quantum Inf. Comput., 13:73, 2013. doi: 10.26421/QIC13.1-2-6.DOI
  23. [23]V. Veitch, S. A. H. Mousavian, D. Gottesman, and J. Emerson. The resource theory of stabilizer quantum computation. New J. Phys., 16:013009, 2014. doi: 10.1088/1367-2630/16/1/013009.DOI
  24. [24]D. Gross. Hudson’s theorem for finite-dimensional quantum systems. J. Math. Phys., 47:122107, 2006. doi: 10.1063/1.2393152.DOI
  25. [25]V. Veitch, C. Ferrie, D. Gross, and J. Emerson. Negative quasi-probability as a resource for quantum computation. New J. Phys., 14:113011, 2012. doi: 10.1088/1367-2630/14/11/113011.DOI
  26. [26]H. Bombin and M. A. Martin-Delgado. Homological error correction: classical and quantum codes. J. Math. Phys., 48:052105, 2007. doi: 10.1063/1.2731356.DOI
  27. [27]P. Webster and S. D. Bartlett. Fault-tolerant quantum gates with defects in topological stabilizer codes. Phys. Rev. A, 102:022403, 2020. doi: 10.1103/PhysRevA.102.022403.DOI
  28. [28]N. Seiberg and W. Zhang. Tori, Klein bottles, and modulo 8 parity/time-reversal anomalies of 2+1d staggered fermions. JHEP, 05:264, 2026. doi: 10.1007/JHEP05(2026)264.
  29. [29]S. Takahashi, Z. Li, and B. Yoshida. A no-go theorem for order-two Clifford electric-magnetic duality, 2026. URL https://arxiv.org/abs/2610.02097v2. Version 2.
  30. [30]X. Yu, X. Chen, A. Roy, and J. C. Y. Teo. Twofold twist defect chain at criticality. Phys. Rev. B, 96:205435, 2017. doi: 10.1103/PhysRevB.96.205435.DOI
  31. [31]W. W. Ho, L. Cincio, H. Moradi, D. Gaiotto, and G. Vidal. Edge-entanglement spectrum correspondence in a nonchiral topological phase and Kramers–Wannier duality. Phys. Rev. B, 91:125119, 2015.DOI
  32. [32]D. Aasen, R. S. K. Mong, and P. Fendley. Topological defects on the lattice: I. The Ising model. J. Phys. A, 49:354001, 2016. doi: 10.1088/1751-8113/49/35/354001.DOI
  33. [33]S. D. Pace, G. Delfino, H. T. Lam, and Ö. M. Aksoy. Gauging modulated symmetries: Kramers–Wannier dualities and non-invertible reflections. SciPost Phys., 18:021, 2025. doi: 10.21468/SciPostPhys.18.1.021.DOI
  34. [34]J. A. Howell. Spans in the module (Z_m)^s. Linear Multilinear Algebra, 19(1):67–77, 1986. doi: 10.1080/03081088808817705.

Paper details

vlnc:W_g6rfkxbm
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 0

downloads · 0

A view counts once per browser per day, a download each time the PDF is requested.

  • Physics and Astronomy

Anthropic Claude, Codex

View statement
Looking for related papers…
Contents