betaFeedback

WorksPhysics and Astronomy

Stabilizer models of achiral abelian topological orders

Abstract

We construct continuous-variable stabilizer codes realizing every achiral 2D abelian topological order, both bosonic and fermionic. Starting from an arbitrary signature-zero integer KK-matrix, we show that the corresponding topological order can be obtained by condensing a lattice of dyons with trivial pairwise mutual braiding in an RN\mathbb{R}^N gauge theory. We implement this condensation microscopically by gauging a non-on-site 1-form symmetry of the RN\mathbb{R}^N toric code.

Introduction

Topological stabilizer codes provide exactly solvable models of abelian topological phases, allowing their universal properties to be understood directly at the microscopic level. The 2D toric code is the paradigmatic example: the fusion rules and braiding statistics of its anyonic excitations can be derived explicitly from the underlying lattice model [1]. At the same time, topological stabilizer codes are a promising avenue toward quantum error correction, since they encode quantum information nonlocally and are therefore intrinsically robust against local errors [2].

In Ref. [3], it was shown that every gappable 2D bosonic abelian topological order—i.e., one that admits a gapped boundary [4]—can be realized by a topological stabilizer code on a system of qudits. For example, the double semion theory [5] admits a stabilizer code realization using Z4\mathbb{Z}_4 qudits [3]. This result has been extended to the fermionic setting in Ref. [6], which constructed Majorana-qudit stabilizer code models of all gappable fermionic abelian topological orders. Conversely, it is believed that 2D qudit or Majorana-qudit stabilizer codes can only realize gappable abelian topological orders; this has been proven in the translation-invariant bosonic setting [7]. Thus, gappability is conjectured to precisely characterize the class of 2D abelian topological orders realizable by stabilizer codes on finite-dimensional local Hilbert spaces.

The authors of Ref. [8] constructed 2D continuous-variable (CV) stabilizer codes [9] realizing certain achiral topological orders that are nevertheless ungappable, such as that of U(1)2×U(1)−4U(1)_2 \times U(1)_{-4} Chern-Simons theory [4]. These examples demonstrate that CV stabilizer codes can realize topological orders beyond those accessible to qudit stabilizer codes, raising a natural question: which 2D topological orders can be realized by CV stabilizer codes? In this paper, we construct stabilizer code realizations of all achiral abelian topological orders, both bosonic and fermionic. We further conjecture that achirality precisely characterizes the class of 2D abelian topological orders realizable by CV stabilizer codes.

Building on the insights of Ref. [8], our construction begins from the observation that every achiral bosonic abelian topological order can be obtained by condensing a 2N2N-dimensional lattice of mutually bosonic dyons in an RN\mathbb{R}^{N} gauge theory. To implement this condensation in a quantum lattice model, we start from the RN\mathbb{R}^{N} toric code and gauge the anomaly-free Z2N\mathbb{Z}^{2N} 1-form symmetry generated by closed string operators of the mutually bosonic dyons. This yields a CV stabilizer code realizing the desired topological order. We further generalize this approach to achiral fermionic topological orders, which can be obtained by condensing fermionic lattices in an RN\mathbb{R}^{N} gauge theory. This requires the introduction of fermionic modes that trivialize the anomaly of the corresponding 1-form symmetry [10]. Gauging the Z2N\mathbb{Z}^{2N} 1-form symmetry then yields a Majorana-CV stabilizer code realizing the achiral fermionic topological order.

Dyon condensation in RN\mathbb{R}^{N} gauge theory

We begin by reviewing the formalism of dyon condensation in RN\mathbb{R}^{N} gauge theory [8], allowing both electric charges and magnetic fluxes as dynamical excitations. The dyons of the theory are labeled by their flux and charge,

(ϕ,q)=(ϕ1,…,ϕN,q1,…,qN)∈R2N,(1)(\boldsymbol{\phi},\mathbf{q}) = (\phi^1,\ldots,\phi^N,q^1,\ldots,q^N) \in\mathbb{R}^{2N}, \tag*{(1)}

with fusion given by addition in R2N\mathbb{R}^{2N}. We use a normalization in which ϕn=1\phi^n=1 represents 2π2\pi flux of the nnth gauge field. The topological spins and mutual braiding statistics are encoded by the symmetric bilinear form

⟨x,y⟩=xTJy,J=(0ININ0).(2)\langle x,y\rangle= x^{T}Jy,\qquad J=\begin{pmatrix}0&I_N\\ I_N&0\end{pmatrix}. \tag*{(2)}

In particular, the topological spin and braiding are

θx=eπi⟨x,x⟩,Bx,y=e2πi⟨x,y⟩,(3)\theta_x=e^{\pi i\langle x,x\rangle},\qquad B_{x,y}=e^{2\pi i\langle x,y\rangle}, \tag*{(3)}

for any pair of dyons x,y∈R2Nx,y\in\mathbb{R}^{2N}. Consider a 2N2N-dimensional lattice of dyons L⊂R2NL\subset\mathbb{R}^{2N}. Assume that the dyons in LL have trivial pairwise braiding, that is,

⟨c,c′⟩∈Z,c,c′∈L.(4)\langle c,c'\rangle\in\mathbb{Z},\qquad c,c'\in L. \tag*{(4)}

In particular, θc=±1\theta_c=\pm1, so every dyon is either a boson or a fermion. We call LL bosonic if every dyon in LL is a boson, and fermionic otherwise.

A bosonic (resp. fermionic) lattice can be condensed in a bosonic (resp. fermionic) system [10, 11]. Condensing LL has two effects. First, dyons that braid nontrivially with any c∈Lc \in L become confined; the dyons that survive as deconfined excitations are those which braid trivially with LL, and therefore form the dual lattice

L∨={a∈R2N:⟨a,c⟩∈Z ∀c∈L}.(5)L^\vee= \{a \in\mathbb{R}^{2N} : \langle a,c\rangle\in\mathbb{Z}\ \forall c \in L\}. \tag*{(5)}

Note that L⊆L∨L \subseteq L^\vee. Second, deconfined dyons that differ by fusion with an element of LL become identified. Thus, the fusion group of the resulting topological order, modulo local excitations, is the discriminant group [12]

A=L∨/L.(6)\mathcal{A} = L^\vee/L. \tag*{(6)}

We illustrate the formalism with four examples obtained by condensing two-dimensional dyon lattices in R\mathbb{R} gauge theory. The condensate lattice LL is generated by c1c_1 and c2c_2, and the dual lattice L∨L^\vee is generated by a1a_1 and a2a_2. These lattices are depicted in Figure 1.

Figure 1. Condensate lattice LL (large dots) and dual lattice L∨L^\vee (small dots) for (a) Z2\mathbb{Z}_2 toric code, (b) double semion, (c) U(1)2×U(1)−4U(1)_2 \times U(1)_{-4}, and (d) the achiral Z5\mathbb{Z}_5 theory.

  1. Z2\mathbb{Z}_2 toric code:

c1=(0,2),a1=(12,0),c2=(1,0),a2=(0,1).\begin{aligned} c_1 &= (0,2), & a_1 &= \left(\frac{1}{2},0\right),\\ c_2 &= (1,0), & a_2 &= (0,1). \end{aligned}

LL is generated by charge 2 and 2π2\pi flux. The unit charge a2a_2 and π\pi flux a1a_1 descend, respectively, to the ee and mm anyons of the toric code. Equivalently, we may take c1=(0,2)c_1 = (0,\sqrt{2}), c2=(2,0)c_2 = (\sqrt{2},0), as shown in Figure 1(a).

  1. Double semion:

c1=(1,1),a1=(12,12),c2=(−1,1),a2=(12,−12).\begin{aligned} c_1 &= (1,1), & a_1 &= \left(\frac{1}{2},\frac{1}{2}\right),\\ c_2 &= (-1,1), & a_2 &= \left(\frac{1}{2},-\frac{1}{2}\right). \end{aligned}

Here, [a1][a_1] and [a2][a_2] are a semion and antisemion, respectively. This condensation can be viewed as a two-step process: first condense (2,0)(2,0) and (0,2)(0,2) to obtain the Z4\mathbb{Z}_4 toric code, then condense its e2m2e^2m^2 boson, corresponding to [c1]=[c2][c_1] = [c_2]. This example admits the following generalization:

  1. U(1)m×U(1)−nU(1)_m \times U(1)_{-n}:

c1=(m2,m2),a1=(12m,12m),c2=(−n2,n2),a2=(12n,−12n).\begin{aligned} c_1 &= \left(\sqrt{\frac{m}{2}},\sqrt{\frac{m}{2}}\right), & a_1 &= \left(\sqrt{\frac{1}{2m}},\sqrt{\frac{1}{2m}}\right),\\ c_2 &= \left(-\sqrt{\frac{n}{2}},\sqrt{\frac{n}{2}}\right), & a_2 &= \left(\sqrt{\frac{1}{2n}},-\sqrt{\frac{1}{2n}}\right). \end{aligned}

The classes [a1][a_1] and [a2][a_2] generate the U(1)mU(1)_m and U(1)−nU(1)_{-n} factors, respectively. The theory is bosonic when both mm and nn are even, and fermionic otherwise.

  1. Achiral Zp\mathbb{Z}_p theory, p≡1(mod4)p \equiv1 \pmod{4}:

c1=(1,1),a1=(p−p2p,p+p2p),c2=(1−p2,1+p2),a2=(pp,−pp).\begin{aligned} c_1 &= (1,1), & a_1 &= \left(\frac{p-\sqrt{p}}{2p},\frac{p+\sqrt{p}}{2p}\right),\\ c_2 &= \left(\frac{1-\sqrt{p}}{2},\frac{1+\sqrt{p}}{2}\right), & a_2 &= \left(\frac{\sqrt{p}}{p},-\frac{\sqrt{p}}{p}\right). \end{aligned}

The condensate lattice is bosonic, with discriminant group A≅Zp\mathcal{A} \cong\mathbb{Z}_p. It is generated by [a1][a_1], with [a2]=−2[a1][a_2] = -2[a_1], and the generator has topological spin

θ[a1]=e2πik/p,k=p−14.(7)\theta_{[a_1]} = e^{2\pi i k/p}, \qquad k = \frac{p-1}{4}. \tag*{(7)}

We note that these examples include minimal representatives of all nontrivial classes in the achiral subgroups of the prime components of the pointed Witt group [13]. Let Wptach⊂Wpt\mathcal{W}_{\mathrm{pt}}^{\mathrm{ach}} \subset\mathcal{W}_{\mathrm{pt}} denote the subgroup represented by achiral abelian bosonic topological orders, and define

Wptach(p):=Wpt(p)∩Wptach≅{Z2,p=2,Z2,p≡1(mod4),0,p≡3(mod4).(8)\mathcal{W}_{\mathrm{pt}}^{\mathrm{ach}}(p) := \mathcal{W}_{\mathrm{pt}}(p) \cap\mathcal{W}_{\mathrm{pt}}^{\mathrm{ach}} \cong \begin{cases} \mathbb{Z}_2, & p=2,\\ \mathbb{Z}_2, & p \equiv1 \pmod{4},\\ 0, & p \equiv3 \pmod{4}. \end{cases} \tag*{(8)}

Minimal representatives of the nontrivial classes are given by U(1)2×U(1)−4U(1)_2 \times U(1)_{-4} for p=2p=2 and the achiral Zp\mathbb{Z}_p theory for each prime p≡1(mod4)p \equiv1 \pmod{4}.

Chern-Simons theory description

Condensing the dyon lattice LL amounts to gauging the 1-form symmetry generated by the corresponding Wilson lines, since gauging has the effect of proliferating symmetry defects [14]—which are precisely the dyons in LL [10, 14]. Here we describe this gauging in the continuum theory. As we will see, this gives rise to a U(1)2NU(1)^{2N} Chern–Simons theory with KK-matrix equal to the Gram matrix of LL:

KIJ=⟨cI,cJ⟩.(9)K_{IJ} = \langle c_I,c_J\rangle. \tag*{(9)}

where

cI=(ϕI,qI),I=1,…,2N,(10)c_I=(\phi_I,q_I),\qquad I=1,\ldots,2N, \tag*{(10)}

is a basis of LL.

First, recall the BF theory [15] description of the topological sector of RN\mathbb{R}^N gauge theory. The theory has NN noncompact gauge fields αI\alpha_I and NN noncompact Lagrange multiplier fields βI\beta_I, with action

SBF=−12π∫βI∧dαI=−14π∫JIJBI∧dBJ.(11)S_{\mathrm{BF}}=-\frac{1}{2\pi}\int\beta_I\wedge d\alpha_I=-\frac{1}{4\pi}\int J_{IJ}B_I\wedge dB_J. \tag*{(11)}

Here we have defined the noncompact R2N\mathbb{R}^{2N} gauge field

B=(β,α)=(β1,…,βN,α1,…,αN).(12)B=(\beta,\alpha)=(\beta_1,\ldots,\beta_N,\alpha_1,\ldots,\alpha_N). \tag*{(12)}

The equations of motion impose the flatness condition dBI=0dB_I=0. The action is invariant under R2N\mathbb{R}^{2N} gauge transformations

BI→BI+dfI.(13)B_I\to B_I+df_I. \tag*{(13)}

Since the gauge fields are noncompact, ff is a genuine R2N\mathbb{R}^{2N}-valued function. Wilson lines are labeled by arbitrary dyons a=(ϕ,q)∈R2Na=(\phi,q)\in\mathbb{R}^{2N} as

Wa(γ)=exp⁡[i∫γaIBI].(14)W_a(\gamma)=\exp\left[i\int_\gamma a_I B_I\right]. \tag*{(14)}

The R2N\mathbb{R}^{2N} 1-form symmetry acts as BI→BI+JIJΛJB_I\to B_I+J_{IJ}\Lambda_J for closed ΛI\Lambda_I; its action on Wilson lines is

Wa(γ)→e2πi⟨a,Γ⟩Wa(γ),(15)W_a(\gamma)\to e^{2\pi i\langle a,\Gamma\rangle}W_a(\gamma), \tag*{(15)}

where ΓI=12π∫γΛI\Gamma_I=\frac{1}{2\pi}\int_\gamma\Lambda_I. Note that ΓI\Gamma_I is unquantized, making this a R2N\mathbb{R}^{2N} 1-form symmetry transformation. We are interested in gauging Z2N\mathbb{Z}^{2N} subgroups of this 1-form symmetry.

Let M=(c1 ⋯ c2N)M=(c_1\ \cdots\ c_{2N}) be the matrix of condensed dyons, such that the condensate lattice L=MZ2NL=M\mathbb{Z}^{2N}. Consider the Z2N\mathbb{Z}^{2N} 1-form subgroup consisting of transformations

BI→BI+JIJMJKλK,(16)B_I\to B_I+J_{IJ}M_{JK}\lambda_K, \tag*{(16)}

where λK\lambda_K are closed (dλK=0d\lambda_K=0) and obey the quantization ∫γλK∈2πZ\int_\gamma\lambda_K\in2\pi\mathbb{Z}. Gauging this 1-form symmetry identifies the gauge field configurations which differ by (16). In other words, these transformations become gauge transformations. Requiring invariance of the Wilson lines Wa(γ)W_a(\gamma) imposes the quantization condition

MTJa∈Z2N.(17)M^TJa\in\mathbb{Z}^{2N}. \tag*{(17)}

We may therefore label these deconfined Wilson lines by an integer vector ℓ=MTJa\ell=M^TJa. Defining A=(JM)−1BA=(JM)^{-1}B, the deconfined Wilson lines are

Wa(γ)=exp⁡[i∫γℓIAI].(18)W_a(\gamma)=\exp\left[i\int_\gamma\ell_I A_I\right]. \tag*{(18)}

Hence AIA_I are properly quantized U(1)U(1) gauge fields. Substituting B=(JM)AB=(JM)A into SBFS_{\mathrm{BF}} gives

S=−14π∫KIJAI∧dAJ.(19)S=-\frac{1}{4\pi}\int K_{IJ}A_I\wedge dA_J. \tag*{(19)}

We thus obtain a U(1)2NU(1)^{2N} Chern–Simons theory with KK-matrix

K=MTJM.(20)K=M^TJM. \tag*{(20)}

The condition (4) follows from the requirement that the theory is properly quantized, KIJ∈ZK_{IJ}\in\mathbb{Z}, and therefore gauge invariant. The theory thus describes a 2D abelian topological order [16]. Moreover, KK is even, meaning that KII∈2ZK_{II}\in2\mathbb{Z} for all II, precisely when LL is bosonic. (When KK is odd, the theory is a spin Chern–Simons theory, whose definition requires a choice of spin structure on spacetime [17].)

The anyons of the theory are labeled by integer vectors ℓ∈Z2N\ell\in\mathbb{Z}^{2N}. Vectors in KZ2NK\mathbb{Z}^{2N} represent local excitations, which are all bosonic when KK is even, but otherwise include fermions. The fusion group is thus

A=Z2N/KZ2N≅L∨/L,(21)\mathcal{A}=\mathbb{Z}^{2N}/K\mathbb{Z}^{2N}\cong L^\vee/L, \tag*{(21)}

in agreement with the dyon condensation picture. To see the second isomorphism, observe that a dyon a∈R2Na\in\mathbb{R}^{2N} belongs to L∨L^\vee precisely when MTJa∈Z2NM^TJa\in\mathbb{Z}^{2N}, and hence L∨=(MTJ)−1Z2NL^\vee=(M^TJ)^{-1}\mathbb{Z}^{2N}. Under the isomorphism L∨→Z2NL^\vee\to\mathbb{Z}^{2N} given by a↦MTJaa\mapsto M^TJa, the sublattice L=MZ2NL=M\mathbb{Z}^{2N} maps to KZ2NK\mathbb{Z}^{2N}. This gives the isomorphism L∨/L≅Z2N/KZ2NL^\vee/L\cong\mathbb{Z}^{2N}/K\mathbb{Z}^{2N}.

The topological spins and mutual braiding statistics are [16, 17]

θℓ=eπiℓTK−1ℓ,Bℓ,ℓ′=e2πiℓTK−1ℓ′.(22)\theta_\ell=e^{\pi i\ell^T K^{-1}\ell},\qquad B_{\ell,\ell'}=e^{2\pi i\ell^T K^{-1}\ell'}. \tag*{(22)}

For odd KK, the topological spin depends on the representative of the equivalence class: attaching a local fermion reverses its sign. These expressions also agree with the dyon condensation picture, since the Gram matrix of L∨L^\vee is K−1K^{-1}. Indeed, defining aI=(MTJ)−1ℓIa_I=(M^TJ)^{-1}\ell_I, where ℓI\ell_I is the IIth basis vector of Z2N\mathbb{Z}^{2N}, gives

(K−1)IJ=⟨aI,aJ⟩.(23)(K^{-1})_{IJ}=\langle a_I,a_J\rangle. \tag*{(23)}

For instance, the lattice LL of the above example 4 gives

K=(2111−p2),K−1=1p(p−1211−2).(24)K=\begin{pmatrix}2&1\\1&1-\frac{p}{2}\end{pmatrix},\qquad K^{-1}=\frac{1}{p}\begin{pmatrix}\frac{p-1}{2}&1\\1&-2\end{pmatrix}. \tag*{(24)}

Under the isomorphism in (21), the anyon [a1]∈L∨/L[a_1]\in L^\vee/L corresponds to [ℓ1]∈Z2N/KZ2N[\ell_1]\in\mathbb{Z}^{2N}/K\mathbb{Z}^{2N}.

Finally, the chiral central charge is given by the signature of KK [18],

c−=sig⁡(K)=n+−n−,(25)c_- = \operatorname{sig}(K) = n_+ - n_-, \tag*{(25)}

where n±n_{\pm} denote the numbers of positive and negative eigenvalues of KK. Since KK is congruent to JJ, it has the same signature by Sylvester’s law of inertia. As sig⁡(J)=0\operatorname{sig}(J) = 0, the resulting topological order is achiral, with c−=0c_- = 0. Conversely, every achiral U(1)2NU(1)^{2N} Chern-Simons theory arises from dyon condensation in RN\mathbb{R}^N gauge theory. To see this, diagonalize KK as

K=OT(D+00−D−)O,(26)K = O^T \begin{pmatrix} D_+ & 0 \\ 0 & -D_- \end{pmatrix} O, \tag*{(26)}

where OO is orthogonal and D±D_{\pm} are positive diagonal N×NN \times N matrices. Define

M=H(D+1/200D−1/2)O,H=12(INININ−IN).(27)M = H \begin{pmatrix} D_+^{1/2} & 0 \\ 0 & D_-^{1/2} \end{pmatrix} O, \qquad H = \frac{1}{\sqrt{2}}\begin{pmatrix} I_N & I_N \\ I_N & -I_N \end{pmatrix}. \tag*{(27)}

Since

HTJH=(IN00−IN),(28)H^T J H = \begin{pmatrix} I_N & 0 \\ 0 & -I_N \end{pmatrix}, \tag*{(28)}

the lattice L=MZ2NL = M\mathbb{Z}^{2N} satisfies K=MTJMK = M^T J M. As every achiral 2D abelian topological order can be described by a zero-signature KK-matrix [16, 17], we conclude that dyon condensation in RN\mathbb{R}^N gauge theory realizes precisely the class of achiral 2D abelian topological orders.

Lattice implementation

We now implement the gauging of the 1-form symmetry on the RN\mathbb{R}^N toric code, which serves as a lattice realization of RN\mathbb{R}^N gauge theory [8].

The RN toric code

The Hilbert space HRN\mathcal{H}_{\mathbb{R}^N} consists of NN continuous-variable modes on each edge of the square lattice. Each mode has position and momentum operators x^\hat{x} and p^\hat{p} satisfying [x^,p^]=i[\hat{x},\hat{p}] = i. We define displacement operators

Zt=e2πitx^,Xs=e−isp^,s,t∈R,(29)Z^t = e^{2\pi i t\hat{x}}, \qquad X^s = e^{-i s\hat{p}}, \qquad s,t \in\mathbb{R}, \tag*{(29)}

which obey the commutation relation

ZtXs=e2πistXsZt.(30)Z^t X^s = e^{2\pi i st} X^s Z^t. \tag*{(30)}

Our normalization is chosen so that ZZ shifts momentum by 2π2\pi, mirroring the convention that ϕ=1\phi= 1 denotes a 2π2\pi flux. We denote the displacement operators acting on the nnth CV mode of edge ee by Xn,esX^s_{n,e} and Zn,etZ^t_{n,e}, with n=1,…,Nn = 1,\ldots,N.

The Hamiltonian of the RN\mathbb{R}^N toric code is

HTC=−∑n,v∫Rdϕ An,vϕ−∑n,p∫Rdq Bn,pq,(31)H_{\mathrm{TC}} = -\sum_{n,v}\int_{\mathbb{R}} d\phi\, A^\phi_{n,v} - \sum_{n,p}\int_{\mathbb{R}} dq\, B^q_{n,p}, \tag*{(31)}

where

An,vn=Znt∣Zn−1 ⁣− ⁣v ⁣− ⁣Zn ⁣− ⁣}Bn,pn=Xn ⁣− ⁣p ⁣− ⁣Xn−1∣∣Xn−1A^n_{n,v} = \left. \begin{array}{c} Z^t_n\\ \vert\\ Z^{-1}_n\!-\!v\!-\!Z_n\!-\! \end{array} \right\} \qquad B^n_{n,p} = \begin{array}{c} X_n\!-\!p\!-\!X_n^{-1}\\ \vert\qquad\vert\\ X_n^{-1} \end{array}

Violations of the vertex constraint are electric gauge charges, whereas violations of the plaquette constraint are magnetic gauge fluxes. In particular, an electric charge qnq_n at vertex vv corresponds to eigenvalue e2πiαqne^{2\pi i\alpha q_n} of An,vαA^\alpha_{n,v} for all α\alpha, and a magnetic flux ϕn\phi_n through plaquette pp corresponds to eigenvalue e2πiβϕne^{2\pi i\beta\phi_n} of Bn,pβB^\beta_{n,p} for all β\beta.

The closed string operators associated with the dyons in LL form a non-on-site Z2N\mathbb{Z}^{2N} 1-form symmetry of the RN\mathbb{R}^N toric code. Let ti∈Z2Nt_i \in\mathbb{Z}^{2N} denote the iith generator, corresponding to the dyon ci=Mti∈Lc_i = Mt_i \in L. There is some freedom in the choice of the microscopic representative of the symmetry transformations associated to tit_i. We consider the symmetry generated by operators

Uc,∂p=Z−ϕ∣X−qZϕ ⁣− ⁣Z−ϕ ⁣− ⁣∣∣X−q ⁣− ⁣p ⁣− ⁣XqZϕU_{c,\partial p} = \begin{array}{c} Z^{-\phi}\\ \vert\\ X^{-q}Z^\phi\!-\!Z^{-\phi}\!-\!\\ \vert\qquad\vert\\ X^{-q}\!-\!p\!-\!X^qZ^\phi \end{array}

where

Zϕ=∏n=1NZnϕn,Xq=∏n=1NXnqn,(32)Z^\phi= \prod_{n=1}^{N} Z_n^{\phi_n}, \qquad X^q = \prod_{n=1}^{N} X_n^{q_n}, \tag*{(32)}

The 1-form symmetry generated by Uc,∂pU_{c,\partial p} is anomaly-free if cic_i is a boson, but anomalous if cic_i is a fermion [10, 14]. We denote the fermion parity pip_i of the dyon cic_i as

pi=⟨ci,ci⟩(mod2).(33)p_i = \langle c_i,c_i\rangle\pmod2. \tag*{(33)}

For each ii such that cic_i is fermionic, we introduce a physical fermionic mode at each vertex vv described by Majorana operators γi,v\gamma_{i,v} and γi,v′\gamma'_{i,v}. The Majorana modes form a Hilbert space Hf\mathcal{H}_f. These ancillary modes trivialize the possible anomalies of the 1-form symmetry [19]. We will gauge the augmented toric code Hamiltonian,

H=HTC−∑i,vFi,vpi,Fi,v=iγi,vγi,v′.(34)H = H_{\mathrm{TC}} - \sum_{i,v} F_{i,v}^{p_i}, \qquad F_{i,v} = i\gamma_{i,v}\gamma'_{i,v}. \tag*{(34)}

Here and below, a fermionic factor raised to pi=0p_i=0 denotes the identity.

Gauging the 1-form symmetry

Adapting the defect-based gauging approach of Ref. [20], our procedure consists of three steps:

  1. Introduce a Z2N\mathbb{Z}_{2N} 2-form gauge field living on the vertices of the lattice, forming a Hilbert space HZ2N\mathcal{H}_{\mathbb{Z}_{2N}} consisting of 2N2N rotor degrees of freedom on each vertex. The total Hilbert space is thus H=HRN⊗Hf⊗HZ2N\mathcal{H}=\mathcal{H}_{\mathbb{R}^{N}}\otimes\mathcal{H}_{f}\otimes\mathcal{H}_{\mathbb{Z}_{2N}}. Each rotor has angular coordinate θ^\hat{\theta} and angular momentum n^\hat{n} satisfying [θ^,n^]=i[\hat{\theta},\hat{n}]=i. The eigenvalues of the 2N2N angular momentum operators on vertex vv encode the symmetry defect gv∈Z2Ng_v\in\mathbb{Z}_{2N} at vv. The Hilbert space HZ2N\mathcal{H}_{\mathbb{Z}_{2N}} is thus spanned by defect configurations ∣{gv}⟩\lvert\{g_v\}\rangle. We define displacement operators

Zt=e2πitn^,Xs=eisθ^,s∈Z, t∈R,(35)Z^t=e^{2\pi i t\hat{n}},\qquad X^s=e^{is\hat{\theta}},\qquad s\in\mathbb{Z},\ t\in\mathbb{R}, \tag*{(35)}

obeying the commutation relation

ZtXs=e2πistXsZt.(36)Z^tX^s=e^{2\pi ist}X^sZ^t. \tag*{(36)}

In particular, Zm=1Z^m=\mathbb{1} for m∈Zm\in\mathbb{Z}. We denote the displacement operators acting on the iith rotor on vertex vv by Zi,vtZ_{i,v}^t and Xi,vsX_{i,v}^s, for i=1,…,2Ni=1,\ldots,2N.

  1. Impose Gauss law constraints Gi,e=1\mathcal{G}_{i,e}=1 for each edge ee, oriented from uu to vv, and i=1,…,2Ni=1,\ldots,2N, where the Gauss law operators satisfy the following properties:

I. Gi,e=Xi,u−1⋅Xi,v⋅Vi,e\mathcal{G}_{i,e}=X_{i,u}^{-1}\cdot X_{i,v}\cdot V_{i,e} for some unitary operator Vi,eV_{i,e} diagonal in the ∣{gv}⟩\lvert\{g_v\}\rangle basis and supported near ee.

II. ⟨{gv=0}∣∏e∈∂pGi,e∣{gv=0}⟩=Uci,∂p\langle\{g_v=0\}\lvert\prod_{e\in\partial p}\mathcal{G}_{i,e}\rvert\{g_v=0\}\rangle=U_{c_i,\partial p} in the bosonic case, and Uci,∂p(Fi,SW(p)Fi,NE(p))piU_{c_i,\partial p}(F_{i,\mathrm{SW}(p)}F_{i,\mathrm{NE}(p)})^{p_i} in general.

III. [Gi,e,Gj,e′]=0[\mathcal{G}_{i,e},\mathcal{G}_{j,e'}]=0.

Taken together, properties I and II imply that Gi,e\mathcal{G}_{i,e} can be interpreted as a short string operator that hops a tit_i symmetry defect from uu to vv. In the fermionic case, the additional Majorana factors in property II implement fermionic gauging by attaching a physical fermion to each fermionic symmetry defect. Property III ensures that the constraints Gi,e=1\mathcal{G}_{i,e}=1 can be imposed simultaneously, either directly or energetically. To obtain a stabilizer code model, we will impose the constraints energetically.

  1. Define the gauged Hamiltonian,

H~=∑{gv}U{gv}HU{gv}−1⊗∣{gv}⟩⟨{gv}∣−∑i,e(Gi,e+Gi,e†),(37)\widetilde{H}=\sum_{\{g_v\}}U_{\{g_v\}}HU_{\{g_v\}}^{-1}\otimes\lvert\{g_v\}\rangle\langle\{g_v\}\rvert-\sum_{i,e}\left(\mathcal{G}_{i,e}+\mathcal{G}_{i,e}^{\dagger}\right), \tag*{(37)}

in terms of defect insertion operators U{gv}U_{\{g_v\}} defined for each defect configuration {gv}\{g_v\}. Specifically,

U{gv}=∏i,v(Ui,γv)gvi,(38)U_{\{g_v\}}=\prod_{i,v}\left(U_{i,\gamma_v}\right)^{g_v^i}, \tag*{(38)}

in terms of elementary defect insertion operators

Ui,γv=⟨{gv′=δv,v′ti}∣∏e∈γvGi,e∣{gv′=0}⟩.(39)U_{i,\gamma_v}=\langle\{g_{v'}=\delta_{v,v'}t_i\}\rvert\prod_{e\in\gamma_v}\mathcal{G}_{i,e}\lvert\{g_{v'}=0\}\rangle. \tag*{(39)}

Here, γv\gamma_v is an arbitrary choice of semi-infinite oriented path terminating at vv. For concreteness, we take γv\gamma_v to be a straight line from the west. Thus Ui,γvU_{i,\gamma_v} creates a tit_i symmetry defect at vertex vv. Note that these operators are well-defined due to commutativity of the Gauss law operators. By construction, H~\widetilde{H} commutes with the Gauss law operators.

We carry out this procedure explicitly by choosing the Gauss law operators

Gi,e=Xeqi ,(40)\mathcal{G}_{i,e}=X_e^{q_i}\,, \tag*{(40)}

where

Qi=Xipi∏j<iZj−qj⋅qi,Qi′=Xi′pi∏j>iZjϕj⋅qi.(41)Q_i=X_{i}^{p_i}\prod_{j<i}Z_j^{-q_j\cdot q_i},\qquad Q_i'=X_{i}^{\prime p_i}\prod_{j>i}Z_j^{\phi_j\cdot q_i}. \tag*{(41)}
Qi′=Xi′pi∏j>iZjϕj⋅qi.(42)Q_i'=X_{i}^{\prime p_i}\prod_{j>i}Z_j^{\phi_j\cdot q_i}. \tag*{(42)}

It is straightforward to verify that these operators satisfy properties I–III. The Majorana dressing is closely related to exact two-dimensional bosonization [22]; products of Majorana hopping operators around a plaquette yield fermion parities at two opposite corners, explaining the parity factors in property II.

The resulting defect insertion operator has the form

Ui,γv=(iγi,v′)pi(∏w∈γvFi,wpi)(∏e∈γvXeqiZNE(e)ϕi),(43)U_{i,\gamma_v}=(i\gamma'_{i,v})^{p_i}\left(\prod_{w\in\gamma_v}F_{i,w}^{p_i}\right)\left(\prod_{e\in\gamma_v}X_e^{q_i}Z_{\mathrm{NE}(e)}^{\phi_i}\right), \tag*{(43)}

where NE(e)\mathrm{NE}(e) denotes the edge situated to the northeast of ee. This operator commutes with An,v′A_{n,v'} unless v=v′v=v', with Bn,pB_{n,p} unless vv is the southwest corner of pp, and with Fi,v′F_{i,v'} unless v=v′v=v' and pi=1p_i=1. Imposing the Gauss law constraints Gi,e=1G_{i,e}=1 energetically, the gauged Hamiltonian is therefore

H~=−∑n,v∫Rdϕ A~ϕ,vn−∑n,p∫Rdq B~q,pn−∑i,vF~i,vΦi−∑i,e(Gi,e+Gi,e′).(44)\widetilde{H}=-\sum_{n,v}\int_{\mathbb{R}}d\phi\,\widetilde{A}^{n}_{\phi,v}-\sum_{n,p}\int_{\mathbb{R}}dq\,\widetilde{B}^{n}_{q,p}-\sum_{i,v}\widetilde{F}^{\Phi_i}_{i,v}-\sum_{i,e}\left(G_{i,e}+G'_{i,e}\right). \tag*{(44)}

where [23]

A~ϕ,vn=−Zn−1∏westZvq⃗ n−Zn∏eastZn,B~n,pn=Xn∏?Xn−1.\widetilde{A}^{n}_{\phi,v}=-Z_n^{-1}\mathop{\prod}_{\text{west}}Z_v^{\vec{q}^{\,n}}-Z_n\mathop{\prod}_{\text{east}}Z_n,\qquad\widetilde{B}^{n}_{n,p}=X_n\mathop{\prod}_{\text{?}}X_n^{-1}.
F~i,v=iγi,vγi,v′Zi,v12.(45)\widetilde{F}_{i,v}=i\gamma_{i,v}\gamma'_{i,v}Z_{i,v}^{\frac{1}{2}}. \tag*{(45)}

Here we have introduced the notation

ZΦ⃗=∏i=12NZiΦi,Zq⃗ n=∏i=12NZiqin.(46)\mathbf{Z}^{\vec{\Phi}}=\prod_{i=1}^{2N}Z_i^{\Phi_i},\qquad\mathbf{Z}^{\vec{q}^{\,n}}=\prod_{i=1}^{2N}Z_i^{q_i^n}. \tag*{(46)}

Topological order of the gauged model

We now examine how the dyon condensation rules arise microscopically in the gauged model. String operators for generic dyons of the original RN\mathbb{R}^N toric code fail to commute with the Gauss law operators along their length, reflecting their confinement. The dyons in L∨L^\vee, however, admit string operators that commute with the Hamiltonian away from their endpoints.

To construct these strings, consider the product

Wa,∂p=∏nA~n,vψnB~n,prn,(47)W_{a,\partial p}=\prod_n\widetilde{A}^{\psi_n}_{n,v}\widetilde{B}^{r_n}_{n,p}, \tag*{(47)}

where a=(ψ,r)a=(\psi,\mathbf{r}) and vv is the southwest corner of pp. Its action on the jjth rotor at vv is Z⟨a,cj⟩Z^{\langle a,c_j\rangle}. For a∈L∨a\in L^\vee, this action is trivial, leaving the closed string

Wa,∂p=−Z−ψX−r ZψX−r Z−ψX−r.(48)W_{a,\partial p}=-Z^{-\psi}X^{-\mathbf{r}} \, Z^\psi X^{-\mathbf{r}} \, Z^{-\psi}X^{-\mathbf{r}}. \tag*{(48)}

Multiplying these elementary loops over a region cancels their action on the interior edges, giving a string supported near the boundary. For a∉L∨a\notin L^\vee, the same construction instead leaves a nontrivial action on rotors throughout the interior. The corresponding open string along an oriented path γ\gamma is

Wa,γ=(∏e∈γZSW⁡(e)η(e)ϵ(e)ψ)(∏e∈γXe−ϵ(e)r).(49)W_{a,\gamma}=\left(\prod_{e\in\gamma}Z_{\operatorname{SW}(e)}^{\eta(e)\epsilon(e)\psi}\right)\left(\prod_{e\in\gamma}X_e^{-\epsilon(e)\mathbf{r}}\right). \tag*{(49)}

Here SW⁡(e)\operatorname{SW}(e) is the edge situated to the southwest of ee, ϵ(e)=+1\epsilon(e)=+1 for eastward or northward steps and −1-1 for westward or southward steps, and η(e)=−1\eta(e)=-1 for horizontal edges and +1+1 for vertical edges. An example is shown in Figure 2(b). Each elementary segment commutes with every Gauss law operator when ⟨a,ci⟩∈Z\langle a,c_i\rangle\in\mathbb{Z} for all ii. The string therefore creates excitations only at its endpoints: a dyon −a-a at the initial endpoint and aa at the final endpoint, with the flux at each endpoint residing on the plaquette southwest of its charge. These are the deconfined anyons predicted by the dyon condensation picture.

Defect insertion and open string operators

Figure 2. (a) The defect insertion operator Ui,γvU_{i,\gamma_v}, which acts as ZΦiZ^{\Phi_i} on solid blue edges and XqiX^{q_i} on solid red edges. The black dot represents vertex vv. The Majorana factors are suppressed. (b) An open string operator Wa,γW_{a,\gamma} for the deconfined dyon a=(ψ,r)∈L∨a=(\psi,\mathbf{r})\in L^\vee, acting as ZψZ^\psi on solid blue edges and X±rX^{\pm\mathbf{r}} on solid/dashed red edges. Wa,γW_{a,\gamma} excites the plaquette and vertex term at each endpoint.

A pair of deconfined anyons a,a′∈L∨a,a'\in L^\vee belong to the same superselection sector, modulo local excitations including physical fermions, whenever a′−a∈La'-a\in L. To see why, consider the local operator

Di,v=X−qion p,X−qion Qi.(50)D_{i,v}=X^{-q_i}\quad\text{on }p,\qquad X^{-q_i}\quad\text{on }Q_i. \tag*{(50)}

The rotor shift in QiQ_i creates charge qiq_i at the southwest corner of pp and flux 2Φi2\Phi_i through pp, while the two CVCV shifts move the charge to vv. Moreover, Di,vD_{i,v} commutes with every Gauss law operator and every F~j,v′\widetilde{F}_{j,v'}. It therefore creates precisely the endpoint excitation of a cic_i string at vv: a flux through the plaquette southwest of a charge. Thus, applying Di,vD_{i,v} at a string endpoint transforms aa into a+cia+c_i, identifying dyon labels modulo LL. We conclude that the anyons of the gauged model have fusion group A=L∨/LA=L^\vee/L, with braiding statistics inherited from the original dyons.

We compute the ground state degeneracy of the purely bosonic model on a torus in Appendix B. We find that GSD=∣A∣\mathrm{GSD}=|A|, with logical operators given by noncontractible strings associated with dyons in L∨L^\vee. Together with the fusion rules and braiding statistics, this calculation confirms that the model realizes the topological order described by the input KK-matrix.

Discussion

We have constructed CV stabilizer code realizations of all achiral 2D abelian topological orders. Our construction implements dyon condensation by gauging a non-on-site 1-form symmetry; a general formalism for gauging anomaly-free 1-form symmetries will be discussed in future work [24]. Conversely, we believe that chiral topological orders cannot be realized by CV stabilizer codes purely in two spatial dimensions. On the other hand, we do expect it is possible to realize any abelian topological order at the surface of a 3D CV stabilizer code with no deconfined bulk excitations. We conjecture that such 3D stabilizer codes can be disentangled by CV Clifford quantum cellular automata (QCAs), which are nontrivial in the Clifford category precisely when the surface topological order is chiral. Related connections between surface topological order and QCAs have been established for qudit systems [25, 26, 27, 28].

The Hamiltonians in this paper should be understood as ideal CV stabilizer models: exact codewords involve non-normalizable states, as in multimode GKP codes [9]. An additional ingredient here is the presence of continuous stabilizers, whose integrals formally impose exact constraints. We expect that these constraints can be softened into finite-strength energy penalties by cutting off the domain of integration. As in Ref. [8], we expect that adding suitable quadratic terms in the CV variables and rotor angular momenta will produce normalizable states with a finite energy gap while preserving the desired topological order.

**Acknowledgments—**We are grateful to Clay Cordova, Michael Levin, and Valerie Wu for insightful discussions. The authors used various versions of Chat-GPT and Claude to assist with drafting and revising exposition, as well as exploring, developing, and checking mathematical arguments. The work of MQ is supported in part by the Simons Collaboration on Ultra-Quantum Matter, which is a grant from the Simons Foundation (Grant No. 651442, D.T.S.). The work of WS is supported by U.S. Department of Energy Grant 5-29073. The authors of this paper were ordered alphabetically.

Stabilizer models of 2 by 2 even KK-matrices

In this appendix, we specialize the construction of the main text to the N=1N=1 bosonic case, then describe two alternative CV stabilizer code constructions of bosonic topological orders described by a 2×22 \times2 even KK-matrix. Let ci=(ϕi,qi)c_i=(\phi_i,q_i), i=1,2i=1,2, generate the condensate lattice L⊂R2L\subset\mathbb{R}^2, with Kij=ϕiqj+qiϕjK_{ij}=\phi_iq_j+q_i\phi_j. The Hilbert space H=HR⊗HZ2\mathcal{H}=\mathcal{H}_{\mathbb{R}}\otimes\mathcal{H}_{\mathbb{Z}_2} is composed of one CV mode per edge and two gauge rotors per vertex.

In this case, the gauged Hamiltonian (44) is

H~=−∑v∫Rdα A~vα−∑p∫Rdβ B~pβ−∑i=1,2∑e(Gi,e+Gi,e†).(51)\widetilde{H}=-\sum_v\int_{\mathbb{R}}d\alpha\,\widetilde{A}^{\alpha}_v-\sum_p\int_{\mathbb{R}}d\beta\,\widetilde{B}^{\beta}_p-\sum_{i=1,2}\sum_e\left(G_{i,e}+G^\dagger_{i,e}\right). \tag*{(51)}

The gauged star and plaquette operators are

A~v=Z−1Z1q1Z2q2Z,B~p=X X−1 Z1ϕ1Z2ϕ2X−1.(52)\widetilde{A}_v=Z^{-1}Z_1^{q_1}Z_2^{q_2}Z,\qquad\widetilde{B}_p=X\,X^{-1}\,Z_1^{\phi_1}Z_2^{\phi_2}X^{-1}. \tag*{(52)}

The Gauss law operators are

G1,e=Xeq1,Zϕ1,G2,e=Xeq2,Zϕ2Q1′Z−ϕ1∣G1,e∣Q1−1Q1′−1Xeq1Q1Q2′Z−ϕ2∣G2,e∣Q2−1Q2′−1Xeq2Q2(53)G_{1,e}=X_e^{q_1},\qquad Z^{\phi_1},\qquad G_{2,e}=X_e^{q_2},\qquad Z^{\phi_2} \quad \begin{array}{c} Q'_1Z^{-\phi_1}\\[-2pt] |\\[-2pt] G_{1,e}\\[-2pt] |\\[-2pt] Q_1^{-1} \end{array} \qquad \begin{array}{c} Q_1'^{-1}X_e^{q_1}Q_1\\[-2pt] \end{array} \qquad \begin{array}{c} Q'_2Z^{-\phi_2}\\[-2pt] |\\[-2pt] G_{2,e}\\[-2pt] |\\[-2pt] Q_2^{-1} \end{array} \qquad \begin{array}{c} Q_2'^{-1}X_e^{q_2}Q_2\\[-2pt] \end{array} \tag*{(53)}

where

Q1=X1,Q1′=X1Z2ϕ2q1,Q2=X2Z1−ϕ1q2,Q2′=X2.(54)Q_1=X_1,\qquad Q'_1=X_1Z_2^{\phi_2q_1},\qquad Q_2=X_2Z_1^{-\phi_1q_2},\qquad Q'_2=X_2. \tag*{(54)}

The first alternative construction uses a simplified choice of Gauss law operators on H=HR⊗HZ2\mathcal{H}=\mathcal{H}_{R}\otimes\mathcal{H}_{\mathbb{Z}_2}. In particular, we define

G1,e′=Xeq1,Zϕ1,G2,e′=Xeq2,Zϕ2X1Z−ϕ1∣G1,e′∣X1−1X1′−1Xeq1X1X2∣G2,e′∣Z−ϕ2X2−1X2′−1Xeq2X2(55)G'_{1,e}=X_e^{q_1},\qquad Z^{\phi_1},\qquad G'_{2,e}=X_e^{q_2},\qquad Z^{\phi_2} \quad \begin{array}{c} X_1Z^{-\phi_1}\\[-2pt] |\\[-2pt] G'_{1,e}\\[-2pt] |\\[-2pt] X_1^{-1} \end{array} \qquad \begin{array}{c} X_1'^{-1}X_e^{q_1}X_1 \end{array} \qquad \begin{array}{c} X_2\\[-2pt] |\\[-2pt] G'_{2,e}\\[-2pt] |\\[-2pt] Z^{-\phi_2}X_2^{-1} \end{array} \qquad \begin{array}{c} X_2'^{-1}X_e^{q_2}X_2 \end{array} \tag*{(55)}

With this choice, the toric code operators AvA_v and BpB_p have gauged counterparts

A~v′=−Z−1Z1q1Z2q2Z,B~p′=X−Z2ϕ2∣∣Z1ϕ1−X−1(56)\widetilde{A}'_v=-Z^{-1}Z_1^{q_1}Z_2^{q_2}Z ,\qquad \widetilde{B}'_p= \begin{array}{c} X-Z_2^{\phi_2}\\[-2pt] | \qquad|\\[-2pt] Z_1^{\phi_1}-X^{-1} \end{array} \tag*{(56)}

The gauged Hamiltonian, with energetically imposed Gauss law constraints, is thus

H~′=−∑v∫Rdα A~v′α−∑p∫Rdβ B~p′β−∑i=1,2∑e(Gi,e′+Gi,e′†).(57)\widetilde{H}'=-\sum_v\int_{\mathbb{R}}d\alpha\,\widetilde{A}'^\alpha_v -\sum_p\int_{\mathbb{R}}d\beta\,\widetilde{B}'^\beta_p -\sum_{i=1,2}\sum_e\left(G'_{i,e}+G_{i,e}^{\prime\dagger}\right). \tag*{(57)}

The second alternative construction implements condensation of c1c_1 and c2c_2 directly in HR\mathcal{H}_R, without introducing gauge rotors. Choose a basis ai=(ψi,ri)a_i=(\psi_i,r_i), i=1,2i=1,2, of the dual lattice L∨L^\vee. The Hamiltonian is composed of elementary hopping operators Ci,eC_{i,e} for the condensed dyons and elementary loop operators Wai,∂pW_{a_i,\partial p} for the surviving dyons:

Hc=−∑i=1,2∑p(Wai,∂p+Wai,∂p†)−∑i=1,2∑e(Ci,e+Ci,e†).(58)H_c=-\sum_{i=1,2}\sum_p\left(W_{a_i,\partial p}+W^\dagger_{a_i,\partial p}\right) -\sum_{i=1,2}\sum_e\left(C_{i,e}+C^\dagger_{i,e}\right). \tag*{(58)}

where

Wai,∂p=Xri∣∣XriZψipX−ri∣∣Z−ψi(59)W_{a_i,\partial p}= \begin{array}{c} X^{r_i}\\[-2pt] | \qquad|\\[-2pt] X^{r_i}Z^{\psi_i}\quad p\quad X^{-r_i}\\[-2pt] | \qquad|\\[-2pt] Z^{-\psi_i} \end{array} \tag*{(59)}

and

C1,e=−Z−ϕ1∣Xeq1Zϕ1 ⁣− ⁣Zϕ1 ⁣− ⁣∣Z−ϕ1,−Z−ϕ1∣Zϕ1 ⁣− ⁣∣Zϕ1,C2,e=−Zϕ2∣Z−ϕ2 ⁣− ⁣∣Xeq2Z−ϕ2,−Zϕ2∣Z−ϕ2 ⁣− ⁣Zϕ2 ⁣− ⁣∣Zϕ2.(60)\begin{aligned} C_{1,e}&=-\begin{array}{c} Z^{-\phi_1}\\[-2pt] \vert\\[-2pt] X_e^{q_1}Z^{\phi_1}\!-\!Z^{\phi_1}\!-\!\\[-2pt] \vert\\[-2pt] Z^{-\phi_1}\end{array},\qquad -\begin{array}{c} Z^{-\phi_1}\\[-2pt] \vert\\[-2pt] Z^{\phi_1}\!-\!\\[-2pt] \vert\\[-2pt] Z^{\phi_1}\end{array},\\ C_{2,e}&=-\begin{array}{c} Z^{\phi_2}\\[-2pt] \vert\\[-2pt] Z^{-\phi_2}\!-\!\\[-2pt] \vert\\[-2pt] X_e^{q_2}Z^{-\phi_2}\end{array},\qquad -\begin{array}{c} Z^{\phi_2}\\[-2pt] \vert\\[-2pt] Z^{-\phi_2}\!-\!Z^{\phi_2}\!-\!\\[-2pt] \vert\\[-2pt] Z^{\phi_2}\end{array}. \tag*{(60)} \end{aligned}

Ground state degeneracy on the torus

In this appendix, we compute the logical operators and ground state degeneracy of the gauged model (44) in the case of bosonic LL, corresponding to a topological order described by an even signature-zero KK matrix. Logical operators are closed string operators for deconfined dyons in L∨L^\vee acting on noncontractible cycles, and closed (contractible and noncontractible) string operators for condensed dyons in LL are stabilizers. Hence

L=H1(X;A)={0X=R2A⊕AX=T2(61)\mathcal{L}=H_1(X;\mathcal{A})= \begin{cases} 0 & X=\mathbb{R}^2\\ \mathcal{A}\oplus\mathcal{A} & X=T^2 \end{cases} \tag*{(61)}

and the ground state degeneracy on the torus is ∣A∣=∣det⁡K∣|\mathcal{A}|=|\det K|. To obtain these results, we use the polynomial formalism for translation-invariant stabilizer codes developed by Haah [29], which we adapt to CV degrees of freedom (real variables and rotors) in App. B 1. We then apply the formalism to our model and derive Eq. (61) in App. B 2.

Polynomial formalism for CVs

We begin with a review of the polynomial formalism for translation-invariant stabilizer codes, appropriately modified to accommodate real variable and rotor degrees of freedom. For an abelian group AA, let RA=A[x±,y±]R_A=A[x^{\pm},y^{\pm}] be the module of Laurent polynomials in variables xx and yy with coefficients in AA. In the dd-dimensional qudit setting, AA is typically taken to be Zd\mathbb{Z}_d. For us, AA will be R\mathbb{R} when considering real variables, and Z\mathbb{Z} or R/Z\mathbb{R}/\mathbb{Z} when considering rotors. Periodic boundary conditions are implemented by the quotient A[x±,y±]/⟨xLx=1,yLy=1⟩A[x^{\pm},y^{\pm}]/\langle x^{L_x}=1,y^{L_y}=1\rangle. We will often abuse notation and use RAR_A for both A[x±,y±]A[x^{\pm},y^{\pm}] as well as its quotient by xLx=1x^{L_x}=1, yLy=1y^{L_y}=1, and the meaning of RAR_A depends on whether the model is on the infinite plane or the torus.

Consider a translation-invariant system with nRn_R real variable degrees of freedom and nZn_Z rotors in each unit cell. The Pauli group of the system is

P=RRnR⊕RZnZ⏟X type operators⊕RRnR⊕R(R/Z)nZ⏟Z type operators.(62)\mathcal{P}=\underbrace{R_R^{n_R}\oplus R_{\mathbb{Z}}^{n_Z}}_{X\text{ type operators}}\oplus\underbrace{R_R^{n_R}\oplus R_{(\mathbb{R}/\mathbb{Z})}^{n_Z}}_{Z\text{ type operators}}. \tag*{(62)}

We adopt the following notation for products of Pauli operators at a single site:

X⃗s⃗=∏i=1nRXisi,Z⃗t⃗=∏i=1nRZiti,X⃗s⃗=∏i=1nZXisi,Z⃗t⃗=∏i=1nZZiti,(63)\vec{X}^{\vec{s}}=\prod_{i=1}^{n_R}X_i^{s_i},\qquad \vec{Z}^{\vec{t}}=\prod_{i=1}^{n_R}Z_i^{t_i},\qquad \vec{X}^{\vec{s}}=\prod_{i=1}^{n_Z}X_i^{s_i},\qquad \vec{Z}^{\vec{t}}=\prod_{i=1}^{n_Z}Z_i^{t_i}, \tag*{(63)}

for si,ti∈Rs_i,t_i\in\mathbb{R}, si∈Zs_i\in\mathbb{Z}, and ti∈R/Zt_i\in\mathbb{R}/\mathbb{Z}. The Xi,ZiX_i,Z_i operators act on the real variable degrees of freedom, while the Xi,ZiX_i,Z_i operators act on the rotor degrees of freedom. A general Pauli operator is a tensor product of X⃗s⃗\vec{X}^{\vec{s}}, Z⃗t⃗\vec{Z}^{\vec{t}}, X⃗s⃗\vec{X}^{\vec{s}}, Z⃗t⃗\vec{Z}^{\vec{t}} operators over different sites r=(rx,ry)\mathbf{r}=(r_x,r_y). Pauli operators are represented as polynomials via

P=⨂rXrs⃗rZrt⃗rXrs⃗rZrt⃗r⟷p=(∑rs⃗rxrxyry∑rs⃗rxrxyry∑rt⃗rxrxyry∑rt⃗rxrxyry).(64)P=\bigotimes_{\mathbf{r}}X_{\mathbf{r}}^{\vec{s}_{\mathbf{r}}}Z_{\mathbf{r}}^{\vec{t}_{\mathbf{r}}}X_{\mathbf{r}}^{\vec{s}_{\mathbf{r}}}Z_{\mathbf{r}}^{\vec{t}_{\mathbf{r}}} \longleftrightarrow p= \begin{pmatrix} \sum_{\mathbf{r}}\vec{s}_{\mathbf{r}}x^{r_x}y^{r_y}\\ \sum_{\mathbf{r}}\vec{s}_{\mathbf{r}}x^{r_x}y^{r_y}\\ \sum_{\mathbf{r}}\vec{t}_{\mathbf{r}}x^{r_x}y^{r_y}\\ \sum_{\mathbf{r}}\vec{t}_{\mathbf{r}}x^{r_x}y^{r_y} \end{pmatrix}. \tag*{(64)}

To distinguish between Pauli operators and their polynomial representation, we use capital letters (PP, SS, etc.) to denote Pauli operators and lowercase letters (pp, ss, etc.) to refer to their polynomial representation. The phase obtained from commuting Pauli operators through one another is

P1P2=e2πiλ(p1,p2)P2P1,λ(p1,p2)=[x0y0](p‾1TΛp2).(65)P_1P_2=e^{2\pi i\lambda(p_1,p_2)}P_2P_1,\qquad\lambda(p_1,p_2)=\left[x^0y^0\right]\left(\overline{p}_1^{T}\Lambda p_2\right). \tag*{(65)}

Here [x0y0]\left[x^0y^0\right] denotes the constant coefficient, taken after imposing the periodic identifications on the torus. The overline denotes the antipode map xrxyry↦x−rxy−ryx^{r_x}y^{r_y}\mapsto x^{-r_x}y^{-r_y} and Λ\Lambda is the symplectic matrix

Λ=(0−1n×n1n×n0),n=nR+nZ.(66)\Lambda=\begin{pmatrix} 0 & -\mathbb{1}_{n\times n} \\ \mathbb{1}_{n\times n} & 0 \end{pmatrix},\qquad n=n_{\mathbb{R}}+n_{\mathbb{Z}}. \tag*{(66)}

A translation-invariant stabilizer code is specified by a translation-invariant, mutually commuting subgroup S⊂P\mathcal{S}\subset\mathcal{P} of the Pauli group that does not contain any nontrivial scalar multiples of the identity. Its code space is the mutual +1+1 eigenspace of all S∈SS\in\mathcal{S}. A stabilizer may be either continuous or discrete: if SS is a continuous stabilizer, then SαS^\alpha for α∈R\alpha\in\mathbb{R} are also stabilizers; if SS is a discrete stabilizer, then SkS^k is a stabilizer for k∈Zk\in\mathbb{Z}.

We may pick a translation-invariant set of generators G\mathcal{G} of these stabilizers. With gRg_{\mathbb{R}} continuous and gZg_{\mathbb{Z}} discrete stabilizer generators per unit cell,

G=RgR⊕ZgZ.(67)\mathcal{G}=\mathbb{R}^{g_{\mathbb{R}}}\oplus\mathbb{Z}^{g_{\mathbb{Z}}}. \tag*{(67)}

The stabilizer subgroup can be thought of as the image of a map σ:G→P\sigma:\mathcal{G}\to\mathcal{P}. This map sends each stabilizer generator to its corresponding Pauli operator; its kernel encodes relations between the stabilizers. The stabilizers commute when

σ†Λσ=0,(68)\sigma^\dagger\Lambda\sigma=0, \tag*{(68)}

where the dagger operation †\dagger indicates transposition followed by the antipode. This condition is understood to mean exactly 00 in the block involving continuous generators, and 00 mod 11 in the block of discrete generators.

An excitation of a stabilizer code is an assignment of a phase to each stabilizer generator. The space of excitations forms E=G^=RgR⊕(R/Z)gZ\mathcal{E}=\widehat{\mathcal{G}}=\mathbb{R}^{g_{\mathbb{R}}}\oplus(\mathbb{R}/\mathbb{Z})^{g_{\mathbb{Z}}}. There is an excitation map ϵ:P→E\epsilon:\mathcal{P}\to\mathcal{E} which maps a Pauli operator to the excitations it creates, defined as

ϵ=σ†Λ.(69)\epsilon=\sigma^\dagger\Lambda. \tag*{(69)}

In this language, the stabilizer condition (B8) can then be written as ϵσ=0\epsilon\sigma=0; any Pauli operator in the image of the stabilizer map creates no nontrivial excitations. We may summarize our discussion so far by saying that the maps σ\sigma and ϵ\epsilon form a chain complex

G→σP→ϵE.(70)\mathcal{G}\xrightarrow{\sigma}\mathcal{P}\xrightarrow{\epsilon}\mathcal{E}. \tag*{(70)}

The logical operators on a stabilizer code are Pauli operators which commute with the stabilizers (and hence create no excitation), which are themselves not stabilizers. The logical group is therefore the homology of this complex at P\mathcal{P}:

L=ker⁡ϵim⁡σ.(71)\mathcal{L}=\frac{\ker\epsilon}{\operatorname{im}\sigma}. \tag*{(71)}

As an illustration, let us analyze the instructive example of the RN\mathbb{R}^{N} toric code (31) on the infinite plane R2\mathbb{R}^{2} and the torus T2T^{2} using this formalism. The purpose of studying this familiar example is to interpret the chain complex of (B10) as a cellular chain complex of the spatial manifold, so that computing the logical group reduces to computing cellular homology. This interpretation is naturally expressed in terms of dyon ribbon operators. The intuition involving ribbon operators generalizes readily to the general bosonic KK-matrix case, which we treat in the next subsection.

The RN\mathbb{R}^{N} toric code has 2N2N R\mathbb{R} degrees of freedom per unit cell (NN on each edge), and consists of 2N2N continuous stabilizer generators per unit cell, given by the An,νA_{n,\nu} and Bn,pB_{n,p} stabilizers for n=1,…,Nn=1,\ldots,N. We therefore have P=R2N⊕R2N\mathcal{P}=\mathbb{R}^{2N}\oplus\mathbb{R}^{2N}, G=R2N\mathcal{G}=\mathbb{R}^{2N}, and E=R2N\mathcal{E}=\mathbb{R}^{2N}. We label plaquettes by the vertex at their southwest corner, and horizontal (vertical) edges by the vertex at their west (south) endpoint. With these conventions, the continuous stabilizers are represented in the polynomial formalism as

v⟷(00(1−xˉ)en(1−yˉ)en),p⟷((y−1)en(1−x)en00).(72)v \longleftrightarrow \begin{pmatrix} 0\\ 0\\ (1-\bar{x})e_n\\ (1-\bar{y})e_n \end{pmatrix}, \qquad p \longleftrightarrow \begin{pmatrix} (y-1)e_n\\ (1-x)e_n\\ 0\\ 0 \end{pmatrix}. \tag*{(72)}

Here ene_n is the NN-component unit vector with 1 in the nn-th component and 0 in all others. The stabilizers can be organized into the stabilizer map σ:G→P\sigma:\mathcal{G}\to\mathcal{P} taking the form

σ=(σAv∣σBp)=(0y−101−x1−xˉ01−yˉ0).(73)\sigma=\left(\sigma_{A_v}\mid\sigma_{B_p}\right) = \begin{pmatrix} 0 & y-1\\ 0 & 1-x\\ 1-\bar{x} & 0\\ 1-\bar{y} & 0 \end{pmatrix}. \tag*{(73)}

where each entry should be understood as a scalar multiple of the N×NN\times N identity matrix. The excitation map ϵ=σ†Λ:P→E\epsilon=\sigma^\dagger\Lambda:\mathcal{P}\to\mathcal{E} is

ϵ=(ϵAvϵBp)=(1−x1−y00001−yˉxˉ−1).(74)\epsilon= \begin{pmatrix} \epsilon_{A_v}\\ \epsilon_{B_p} \end{pmatrix} = \begin{pmatrix} 1-x & 1-y & 0 & 0\\ 0 & 0 & 1-\bar{y} & \bar{x}-1 \end{pmatrix}. \tag*{(74)}

We now explain the geometric picture that will allow us to compute the logical group L=ker⁡ϵ/im⁡σ\mathcal{L}=\ker\epsilon/\operatorname{im}\sigma as the cellular homology group of the spatial manifold R2\mathbb{R}^{2} or T2\mathbb{T}^{2}. The cellular chain groups C2(X;R2N)C_2(X;\mathbb{R}^{2N}), C1(X;R2N)C_1(X;\mathbb{R}^{2N}), and C0(X;R2N)C_0(X;\mathbb{R}^{2N}) are formal sums with R2N\mathbb{R}^{2N} coefficients of faces, edges, and vertices, respectively. They can be described in the polynomial language as

C2(X;R2N)=RR2N,C1(X;R2N)=RR2N⊕RR2N,C0(X;R2N)=RR2N.(75)C_2(X;\mathbb{R}^{2N})=R_{\mathbb{R}^{2N}},\qquad C_1(X;\mathbb{R}^{2N})=R_{\mathbb{R}^{2N}}\oplus R_{\mathbb{R}^{2N}},\qquad C_0(X;\mathbb{R}^{2N})=R_{\mathbb{R}^{2N}}. \tag*{(75)}

With the same labeling conventions for faces and edges as before, the boundary maps ∂2:C2(X;R2N)→C1(X;R2N)\partial_2:C_2(X;\mathbb{R}^{2N})\to C_1(X;\mathbb{R}^{2N}) and ∂1:C1(X;R2N)→C0(X;R2N)\partial_1:C_1(X;\mathbb{R}^{2N})\to C_0(X;\mathbb{R}^{2N}) are given by

∂2=(1−yx−1),∂1=(x−1y−1).(76)\partial_2= \begin{pmatrix} 1-y\\ x-1 \end{pmatrix}, \qquad \partial_1= \begin{pmatrix} x-1 & y-1 \end{pmatrix}. \tag*{(76)}

Each entry above should be understood as a scalar multiple of the 2N×2N2N\times2N identity matrix. The homology group H1(X;R2N)=ker⁡∂1/im⁡∂2H_1(X;\mathbb{R}^{2N})=\ker\partial_1/\operatorname{im}\partial_2 vanishes for X=R2X=\mathbb{R}^{2} and is R2N⊕R2N\mathbb{R}^{2N}\oplus\mathbb{R}^{2N} for X=T2X=\mathbb{T}^{2}.

Notice that, as abelian groups, G≅C2(X;R2N)\mathcal{G}\cong C_2(X;\mathbb{R}^{2N}), P≅C1(X;R2N)\mathcal{P}\cong C_1(X;\mathbb{R}^{2N}), and E≅C0(X;R2N)\mathcal{E}\cong C_0(X;\mathbb{R}^{2N}). Importantly, these isomorphisms can be chosen so that σ\sigma and ϵ\epsilon become the boundary maps ∂2\partial_2 and ∂1\partial_1. We identify G≅C2(X;R2N)\mathcal{G}\cong C_2(X;\mathbb{R}^{2N}) by the map (α,β)↦(α,β)(\alpha,\beta)\mapsto(\alpha,\beta) in RR2NR_{\mathbb{R}^{2N}}. This assigns the plaquette stabilizer to its corresponding face, and the vertex stabilizer to its northeast face. We identify P≅C1(X;R2N)\mathcal{P}\cong C_1(X;\mathbb{R}^{2N}) using ribbon operators. To an edge ee with coefficient (ψ,r)(\psi,r), we assign the Pauli operator

Xe−r∣∣Z−ψXe−r∣∣Zψ(77)\begin{array}{c} X_e^{-r}\\[-2pt] \vert\qquad\vert\\[-2pt] Z^{-\psi}\qquad\qquad X_e^{-r}\\[-2pt] \vert\qquad\vert\\[-2pt] \qquad Z^\psi \tag*{(77)} \end{array}

which is the elementary ribbon operator Wa,eW_{a,e} (see Eq. (49)) for the dyon a=(ψ,r)a=(\psi,r) associated to the edge $e. It creates the dyon −a-a at the west (south) endpoint of ee and aa at the east (north), with each flux in the face southwest of the charge. Every Pauli operator on the real variable degrees of freedom can be uniquely written as a product of these elementary ribbon operators, which identifies P\mathcal{P} with 1-chains:

P∋p=(rhrvψhψv)⟷(−yψv−rhxψh−rv)∈RR2N⊕RR2N≅C1(X;R2N).(78)\mathcal{P} \ni p = \begin{pmatrix} r_{\mathrm{h}} \\ r_{\mathrm{v}} \\ \psi_{\mathrm{h}} \\ \psi_{\mathrm{v}} \end{pmatrix} \longleftrightarrow \begin{pmatrix} -y\psi_{\mathrm{v}} \\ -r_{\mathrm{h}} \\ x\psi_{\mathrm{h}} \\ -r_{\mathrm{v}} \end{pmatrix} \in\mathbb{R}_{\mathbb{R}^{2N}} \oplus\mathbb{R}_{\mathbb{R}^{2N}} \cong C_1(X;\mathbb{R}^{2N}) . \tag*{(78)}

Since the elementary ribbon operator excites only its endpoints, we identify E≅C0(X;R2N)\mathcal{E} \cong C_0(X;\mathbb{R}^{2N}) by associating the charge and flux of each endpoint excitation to its corresponding endpoint vertex. This gives the isomorphism

E∋e=(qϕ)⟷(xyϕq)∈RR2N≅C0(X;R2N).(79)\mathcal{E} \ni e = \begin{pmatrix} q \\ \phi \end{pmatrix} \longleftrightarrow \begin{pmatrix} x y \phi\\ q \end{pmatrix} \in\mathbb{R}_{\mathbb{R}^{2N}} \cong C_0(X;\mathbb{R}^{2N}) . \tag*{(79)}

Under this identification, the flux excitation at a plaquette lives at the vertex at its northeast corner. One can check that, using the identifications above,

P∋σ(αβ)⟷∂2(αβ)∈C1(X;R2N)(80)\mathcal{P} \ni\sigma \begin{pmatrix} \alpha\\ \beta \end{pmatrix} \longleftrightarrow \partial_2 \begin{pmatrix} \alpha\\ \beta \end{pmatrix} \in C_1(X;\mathbb{R}^{2N}) \tag*{(80)}

and

E∋ϵ(rhrvψhψv)⟷∂1(−yψv−rhxψh−rv)∈C0(X;R2N).(81)\mathcal{E} \ni\epsilon \begin{pmatrix} r_{\mathrm{h}} \\ r_{\mathrm{v}} \\ \psi_{\mathrm{h}} \\ \psi_{\mathrm{v}} \end{pmatrix} \longleftrightarrow \partial_1 \begin{pmatrix} -y\psi_{\mathrm{v}} \\ -r_{\mathrm{h}} \\ x\psi_{\mathrm{h}} \\ -r_{\mathrm{v}} \end{pmatrix} \in C_0(X;\mathbb{R}^{2N}) . \tag*{(81)}

We therefore identify σ\sigma with ∂2\partial_2, and ϵ\epsilon with ∂1\partial_1. Thus the logical group is L=ker⁡ϵ/im⁡σ=ker⁡∂1/im⁡∂2=H1(X;R2N)\mathcal{L} = \ker\epsilon/ \operatorname{im}\sigma= \ker\partial_1 / \operatorname{im}\partial_2 = H_1(X;\mathbb{R}^{2N}). On the torus, the logical group is continuous, and the code space is infinite-dimensional. The number of encoded real variable degrees of freedom follows from a generalization of the usual counting argument: on an Lx×LyL_x \times L_y torus, there are 2NLxLy2N L_x L_y R\mathbb{R} degrees of freedom, and 2NLxLy−2N2N L_x L_y - 2N independent continuous stabilizers which each remove one R\mathbb{R} degree of freedom. Thus the RN\mathbb{R}^N toric code encodes 2N2N R\mathbb{R} degrees of freedom.

Application to the lattice model

We will now use the polynomial approach to analyze the Hamiltonian (44) in the bosonic case, and show that it realizes the topological order described by the corresponding KK-matrix Chern-Simons theory. In particular, we show that the group of logical operators on the torus is

L(T2)≅H1(T2;A)=A⊕A,A=Z2N/KZ2N.(82)\mathcal{L}(T^2) \cong H_1(T^2;\mathcal{A}) = \mathcal{A} \oplus\mathcal{A}, \qquad\mathcal{A} = \mathbb{Z}^{2N}/K\mathbb{Z}^{2N}. \tag*{(82)}

Our method will be similar in spirit to the one used above in the analysis of the RN\mathbb{R}^N toric code: we will show that the stabilizer and excitation maps are related to boundary maps of an appropriate cellular chain complex. The additional ingredients are the discrete Gauss law stabilizers, which have the effect of confining dyons outside of L∨L^\vee, and turning closed (contractible or noncontractible) loops of ribbon operators of LL into stabilizers.

The Hilbert space of our model consists of NN real variable degrees of freedom on each edge, and 2N2N rotors on each site, for a total of 4N4N CV degrees of freedom per unit cell. The Pauli group and symplectic form are

P=RR2N⊕RZ2N⊕RR2N⊕R(R/Z)2N,Λ=(0−14N×4N14N×4N0).(83)\mathcal{P} = \mathbb{R}_{\mathbb{R}^{2N}} \oplus\mathbb{R}_{\mathbb{Z}^{2N}} \oplus\mathbb{R}_{\mathbb{R}^{2N}} \oplus\mathbb{R}_{(\mathbb{R}/\mathbb{Z})^{2N}}, \qquad \Lambda= \begin{pmatrix} 0 & -\mathbb{1}_{4N \times4N} \\ \mathbb{1}_{4N \times4N} & 0 \end{pmatrix}. \tag*{(83)}

Let us also denote by Posc\mathcal{P}_{\mathrm{osc}} the subgroup of Pauli operators which only act on the R\mathbb{R} degrees of freedom

Posc=RR2N⊕RR2N.(84)\mathcal{P}_{\mathrm{osc}} = \mathbb{R}_{\mathbb{R}^{2N}} \oplus\mathbb{R}_{\mathbb{R}^{2N}} . \tag*{(84)}

The Hamiltonian H~\widetilde{H} consists of 6N6N stabilizer generators per unit cell: NN star terms, NN plaquette terms, 2N2N horizontal Gauss laws, and 2N2N vertical Gauss laws. The star and plaquette stabilizers are continuous stabilizers, while the Gauss law stabilizers are discrete stabilizers. The space of stabilizer generators is therefore

G=GAB⊕GG=(RRN⊕RRN)⊕(RZ2N⊕RZ2N).(85)\mathcal{G}=\mathcal{G}_{AB}\oplus\mathcal{G}_{G}=\left(\mathbb{R}_{\mathbb{R}N}\oplus\mathbb{R}_{\mathbb{R}N}\right)\oplus\left(\mathbb{R}_{\mathbb{Z}_{2}N}\oplus\mathbb{R}_{\mathbb{Z}_{2}N}\right). \tag*{(85)}

The space of excitations is

E=G^=EAB⊕EG=(RRN⊕RRN)⊕(R(R/Z)2N⊕R(R/Z)2N).(86)\mathcal{E}=\widehat{\mathcal{G}}=\mathcal{E}_{AB}\oplus\mathcal{E}_{G}=\left(\mathbb{R}_{\mathbb{R}N}\oplus\mathbb{R}_{\mathbb{R}N}\right)\oplus\left(\mathbb{R}_{(\mathbb{R}/\mathbb{Z})2N}\oplus\mathbb{R}_{(\mathbb{R}/\mathbb{Z})2N}\right). \tag*{(86)}

The stabilizer map consists of 8N8N rows and 6N6N columns, and takes the form

σ=(σAvσBpσGhσGv)=(000(y−1)1N×NQT0000(1−x)1N×N0QT0000(x−1)12N×2N(y−1)12N×2N(1−xˉ)1N×N0000−yΦ‾T(1−yˉ)1N×N000xΦ‾T0QΦ−QhQv).(87)\begin{aligned} \sigma=\left(\begin{array}{c|c|c|c} \sigma_{A_v}&\sigma_{B_p}&\sigma_{G_h}&\sigma_{G_v} \end{array}\right) \\ &=\left(\begin{array}{ccc|c|c|c} 0&0&0&(y-1)\mathbb{1}_{N\times N}&Q^{T}&0\\ 0&0&0&(1-x)\mathbb{1}_{N\times N}&0&Q^{T}\\ 0&0&0&0&(x-1)\mathbb{1}_{2N\times2N}&(y-1)\mathbb{1}_{2N\times2N}\\ (1-\bar{x})\mathbb{1}_{N\times N}&0&0&0&0&-y\overline{\Phi}^{T}\\ (1-\bar{y})\mathbb{1}_{N\times N}&0&0&0&x\overline{\Phi}^{T}&0\\ Q&\Phi&-Q_h& & &Q_v \end{array}\right). \tag*{(87)} \end{aligned}

We have defined the matrices

Q=(q1q2⋮q2N),Φ=(ϕ1ϕ2⋮ϕ2N),Qh=(ΦQT)L.T.+x(ΦQT)U.T.,Qv=y(ΦQT)L.T.+(ΦQT)U.T..(88)Q=\begin{pmatrix}q_{1}\\q_{2}\\\vdots\\q_{2N}\end{pmatrix},\qquad \Phi=\begin{pmatrix}\phi_{1}\\\phi_{2}\\\vdots\\\phi_{2N}\end{pmatrix},\qquad Q_h=(\Phi Q^{T})_{L.T.}+x(\Phi Q^{T})_{U.T.},\qquad Q_v=y(\Phi Q^{T})_{L.T.}+(\Phi Q^{T})_{U.T.}. \tag*{(88)}

where L.T.L.T. and U.T.U.T. signify the strictly lower triangular and upper triangular parts of the matrix ΦQT\Phi Q^{T}. The excitation map ϵ=σ†Λ\epsilon=\sigma^{\dagger}\Lambda is

ϵ=(ϵAϵBϵhGϵvG)=((1−x)1N×N(1−y)1N×NQT00000ΦT(1−yˉ)1N×N(xˉ−1)1N×N00xˉΦ−Q‾hT−Q0(1−xˉ)12N×2N−yˉΦ0Q‾vT0−Q(1−yˉ)12N×2N).(89)\epsilon= \begin{pmatrix} \epsilon_A\\ \epsilon_B\\ \epsilon_h^G\\ \epsilon_v^G \end{pmatrix} = \begin{pmatrix} (1-x)\mathbb{1}_{N\times N}&(1-y)\mathbb{1}_{N\times N}&Q^{T}&0&0&0\\ 0&0&\Phi^{T}&(1-\bar{y})\mathbb{1}_{N\times N}&(\bar{x}-1)\mathbb{1}_{N\times N}&0\\ 0&\bar{x}\Phi&-\overline{Q}_h^{T}&-Q&0&(1-\bar{x})\mathbb{1}_{2N\times2N}\\ -\bar{y}\Phi&0&\overline{Q}_v^{T}&0&-Q&(1-\bar{y})\mathbb{1}_{2N\times2N} \end{pmatrix}. \tag*{(89)}

It can be verified explicitly that ϵσ=0\epsilon\sigma=0, with the Gauss law excitation rows understood modulo integers. Our aim now is to compute the group of logical operators L\mathcal{L} on the infinite plane and the torus. Recall that L\mathcal{L} is given by the homology of the chain complex (B10) at P\mathcal{P}. We will compute this homology group in two steps. First, we show that the support of any operator that creates no excitation (equivalently, any operator in ker⁡ϵ\ker\epsilon) can be “cleaned” so that it only acts on the real variable degrees of freedom and commutes with the Gauss laws. This allows us to consider instead the complex

Gosc→σPosc∩ker⁡ϵG→ϵABEAB.(90)\mathcal{G}_{\mathrm{osc}}\xrightarrow{\sigma}\mathcal{P}_{\mathrm{osc}}\cap\ker\epsilon_G\xrightarrow{\epsilon_{AB}}\mathcal{E}_{AB}. \tag*{(90)}

Here, we have defined Gosc=σ−1(Posc)\mathcal{G}_{\mathrm{osc}}=\sigma^{-1}(\mathcal{P}_{\mathrm{osc}}) as the subspace of G\mathcal{G} which maps to stabilizers supported on the real variable degrees of freedom, and EAB\mathcal{E}_{AB} as the subspace of excitations which violate only the star and plaquette stabilizers. Next, we will show that this chain complex admits a simple geometric interpretation in terms of homology on the plane and the torus. We show, for the square lattice cellulation of X=R2X=\mathbb{R}^{2} and X=T2X=T^{2}, that

Gosc≅C2(X;L∨)⊕Z1(X;L),Posc∩ker⁡ϵG≅C1(X;L∨),EAB≅C0(X;R2N).(91)\mathcal{G}_{\mathrm{osc}}\cong C_{2}(X;L^{\vee})\oplus Z_{1}(X;L),\qquad \mathcal{P}_{\mathrm{osc}}\cap\ker\epsilon_G\cong C_{1}(X;L^{\vee}),\qquad \mathcal{E}_{AB}\cong C_{0}(X;\mathbb{R}^{2N}). \tag*{(91)}

and provide identifications such that σ\sigma and ϵAB\epsilon_{AB} can be interpreted as geometric boundary maps. It then follows that the homology of this complex is canonically isomorphic to H1(X;A)H_{1}(X;A). This verifies the natural expectation that there are no nontrivial logical operators on the plane, and that the logical operators on the torus are represented by string operators supported on noncontractible cycles of the torus.

The first step is to show that any element of ker⁡ϵ\ker\epsilon has a representative in Posc\mathcal{P}_{\mathrm{osc}}. Let

p=(rhrvsψhψvt),rh,rv,ψh,ψv∈RRN,s∈RZ2N,t∈R(R/Z)2N,(92)p=\begin{pmatrix} r_h \\ r_v \\ \mathbf{s} \\ \psi_h \\ \psi_v \\ \mathbf{t} \end{pmatrix},\qquad r_h,r_v,\psi_h,\psi_v\in R_{\mathbb{R}^N},\quad\mathbf{s}\in R_{\mathbb{Z}^{2N}},\quad\mathbf{t}\in R_{(\mathbb{R}/\mathbb{Z})^{2N}}, \tag*{(92)}

be an arbitrary element of ker⁡ϵ\ker\epsilon. The first 2N2N rows of ϵp=0\epsilon p=0 are

(1−x)rh+(1−y)rv+QTs=0(1-x)r_h+(1-y)r_v+Q^T\mathbf{s}=0
(1−yˉ)ψh+(xˉ−1)ψv+ΦTs=0.(93)(1-\bar{y})\psi_h+(\bar{x}-1)\psi_v+\Phi^T\mathbf{s}=0. \tag*{(93)}

Evaluating both sides at x=y=1x=y=1 (in other words, summing over all coefficients) gives

(QTΦT)s∣x=y=1=0.(94)\left.\begin{pmatrix} Q^T \\ \Phi^T \end{pmatrix}\mathbf{s}\right|_{x=y=1}=0. \tag*{(94)}

Since ∣det⁡(QTΦT)∣=∣A∣≠0\left|\det\begin{pmatrix} Q^T \\ \Phi^T \end{pmatrix}\right|=\sqrt{|\mathcal{A}|}\ne0, we have s∣x=y=1=0\left.\mathbf{s}\right|_{x=y=1}=0. This means that s\mathbf{s} can be written as

s=(1−x)h+(1−y)v,h,v∈RZ2N.(95)\mathbf{s}=(1-x)\mathbf{h}+(1-y)\mathbf{v},\qquad\mathbf{h},\mathbf{v}\in R_{\mathbb{Z}^{2N}}. \tag*{(95)}

We can then add Gauss law stabilizers

p′=p+σ(00hv)(96)p'=p+\sigma\begin{pmatrix} 0 \\ 0 \\ \mathbf{h} \\ \mathbf{v} \end{pmatrix} \tag*{(96)}

to pp so that the RZ2NR_{\mathbb{Z}^{2N}} component of p′p' vanishes. Having shown this, let us now take an arbitrary element p∈ker⁡ϵp\in\ker\epsilon with vanishing s\mathbf{s}, but not necessarily vanishing t\mathbf{t}. To this pp, we can add AvA_v and BpB_p stabilizers

p′=p+σ(αβ00),α,β∈RRN(97)p'=p+\sigma\begin{pmatrix} \alpha\\ \beta\\ 0 \\ 0 \end{pmatrix},\qquad\alpha,\beta\in R_{\mathbb{R}^N} \tag*{(97)}

for α,β\alpha,\beta satisfying

Qα+Φβ=(Q Φ)(αβ)=−t.(98)Q\alpha+\Phi\beta=(Q\ \Phi)\begin{pmatrix}\alpha\\\beta\end{pmatrix}=-\mathbf{t}. \tag*{(98)}

Here, in an abuse of notation, we have used the same symbol t\mathbf{t} to denote its lift into RR2NR_{\mathbb{R}^{2N}}. For a given lift, there is a unique α,β\alpha,\beta satisfying the above since ∣det⁡((Q Φ))∣=∣A∣≠0\left|\det\left((Q\ \Phi)\right)\right|=\sqrt{|\mathcal{A}|}\ne0. After adding this combination of star and plaquette stabilizers, p′p' acts nontrivially only on the real variable degrees of freedom. This allows us to restrict to the subcomplex

Gosc→ σ Posc→ ϵ E(99)\mathcal{G}_{\mathrm{osc}}\xrightarrow{\ \sigma\ }\mathcal{P}_{\mathrm{osc}}\xrightarrow{\ \epsilon\ }\mathcal{E} \tag*{(99)}

without changing the homology ker⁡ϵ/im⁡σ\ker\epsilon/\operatorname{im}\sigma. Finally, writing ϵ=(ϵABϵG)\epsilon=\begin{pmatrix}\epsilon_{AB}\\\epsilon_G\end{pmatrix}, we can further restrict Posc\mathcal{P}_{\mathrm{osc}} to ker⁡ϵG\ker\epsilon_G to get (B30). Since logical operators belong to the kernel of ϵ\epsilon, we lose nothing by making this further restriction to ker⁡ϵG\ker\epsilon_G. This also allows us to restrict E\mathcal{E} to EAB\mathcal{E}_{AB}, the subspace of excitations of the vertex and plaquette stabilizers. This completes the reduction to the chain complex (B30).

We have shown that the logical operators are given by the homology of (B30) at the middle term Posc∩ker⁡ϵG\mathcal{P}_{\mathrm{osc}} \cap\ker\epsilon_G. We now provide the geometric interpretation of the chain complex in terms of cellular homology, beginning with the map ϵAB:Posc∩ker⁡ϵG→EAB\epsilon_{AB} : \mathcal{P}_{\mathrm{osc}} \cap\ker\epsilon_G \to\mathcal{E}_{AB}. Recall from our analysis of the RN\mathbb{R}^N toric code that

Posc≅C1(X;R2N)(100)\mathcal{P}_{\mathrm{osc}} \cong C_1(X;\mathbb{R}^{2N}) \tag*{(100)}

where a ribbon operator (B17) in Posc\mathcal{P}_{\mathrm{osc}} is mapped to its corresponding edge in C1(X;R2N)C_1(X;\mathbb{R}^{2N}) via (B18). Similarly, EAB\mathcal{E}_{AB} is the set of excitations of the AvA_v and BpB_p stabilizers. Just as in the RN\mathbb{R}^N toric code, we associate to a vertex vv with coefficient (ϕ,q)∈R2N(\phi,\mathbf{q}) \in\mathbb{R}^{2N} the dyon excitation with q\mathbf{q} at vv and ϕ\phi at the plaquette southwest of vv, giving EAB≅C0(X;R2N)\mathcal{E}_{AB} \cong C_0(X;\mathbb{R}^{2N}).

Restricting Posc\mathcal{P}_{\mathrm{osc}} to Posc∩ker⁡ϵG\mathcal{P}_{\mathrm{osc}} \cap\ker\epsilon_G restricts the coefficients of C1(X;R2N)C_1(X;\mathbb{R}^{2N}) to C1(X;L∨)C_1(X;L^\vee). To see this, let p∈Poscp \in\mathcal{P}_{\mathrm{osc}} be given by (B32), with s=t=0\mathbf{s}=\mathbf{t}=0. Elements of Posc∩ker⁡ϵG\mathcal{P}_{\mathrm{osc}} \cap\ker\epsilon_G obey

xˉΦrv−Qψh=0(mod1)\bar{x}\Phi\mathbf{r}_v-Q\boldsymbol{\psi}_h=0 \pmod1
−yˉΦrh−Qψv=0(mod1)(101)-\bar{y}\Phi\mathbf{r}_h-Q\boldsymbol{\psi}_v=0 \pmod1 \tag*{(101)}

These conditions imply that the ribbon operators associated to

(−yψv−rh),(xψh−rv)\begin{pmatrix} -y\boldsymbol{\psi}_v\\ -\mathbf{r}_h \end{pmatrix}, \begin{pmatrix} x\boldsymbol{\psi}_h\\ -\mathbf{r}_v \end{pmatrix}

create dyons in L∨⊂R2NL^\vee\subset\mathbb{R}^{2N}. Therefore Posc∩ker⁡ϵG≅C1(X;L∨)\mathcal{P}_{\mathrm{osc}} \cap\ker\epsilon_G \cong C_1(X;L^\vee). The map ϵAB\epsilon_{AB} acts on Posc∩ker⁡ϵG\mathcal{P}_{\mathrm{osc}} \cap\ker\epsilon_G in precisely the same way as the excitation map ϵ\epsilon of the RN\mathbb{R}^N toric code. We can therefore identify ϵAB\epsilon_{AB} with ∂1\partial_1.

Let us now understand Gosc\mathcal{G}_{\mathrm{osc}} and σ\sigma. First, observe that G\mathcal{G} is isomorphic to C2(X;R2N)⊕C1(X;Z2N)C_2(X;\mathbb{R}^{2N}) \oplus C_1(X;\mathbb{Z}^{2N}): given an element g=(α β h v)T∈Gg=(\alpha\ \beta\ \mathbf{h}\ \mathbf{v})^{T}\in\mathcal{G}, we can think of

(αβ)\begin{pmatrix} \alpha\\ \beta \end{pmatrix}

as an element of C2(X;R2N)C_2(X;\mathbb{R}^{2N}) and

(hv)\begin{pmatrix} \mathbf{h}\\ \mathbf{v} \end{pmatrix}

as an element of C1(X;Z2N)C_1(X;\mathbb{Z}^{2N}). For gg to be an element of Gosc\mathcal{G}_{\mathrm{osc}}, σ(g)\sigma(g) must have vanishing support on the rotor degrees of freedom. This requirement imposes

(x−1)h+(y−1)v=0(x-1)\mathbf{h}+(y-1)\mathbf{v}=0
Qα+Φβ−Qhh+Qvv=0(mod1)(102)Q\alpha+\Phi\beta-Q\mathbf{h}\mathbf{h}+Q\mathbf{v}\mathbf{v}=0 \pmod1 \tag*{(102)}

The first equation above states that (hv)∈C1(X;Z2N)\begin{pmatrix}\mathbf{h}\\\mathbf{v}\end{pmatrix}\in C_1(X;\mathbb{Z}^{2N}) is closed, so (hv)∈Z1(X;Z2N)\begin{pmatrix}\mathbf{h}\\\mathbf{v}\end{pmatrix}\in Z_1(X;\mathbb{Z}^{2N}). To make sense of the second equation, define α~≡α−ΦT(v−xh)\widetilde{\alpha}\equiv\alpha-\Phi^{T}(\mathbf{v}-x\mathbf{h}). The second equation becomes

Qα~+Φβ=0(mod1).(103)Q\widetilde{\alpha}+\Phi\beta=0 \pmod1 . \tag*{(103)}

This step makes use of the fact that (hv)\begin{pmatrix}\mathbf{h}\\\mathbf{v}\end{pmatrix} is closed. Therefore, (α~β)\begin{pmatrix}\widetilde{\alpha}\\\beta\end{pmatrix} is an element of C2(X;L∨)⊂C2(X;R2N)C_2(X;L^\vee)\subset C_2(X;\mathbb{R}^{2N}). This gives the isomorphism

Gosc∋(αβhv)⟷((α−ΦT(v−xh)β),(hv))∈C2(X;L∨)⊕Z1(X;Z2N).(104)\mathcal{G}_{\mathrm{osc}}\ni \begin{pmatrix} \alpha\\ \beta\\ \mathbf{h}\\ \mathbf{v} \end{pmatrix} \longleftrightarrow \left( \begin{pmatrix} \alpha-\Phi^{T}(\mathbf{v}-x\mathbf{h})\\ \beta \end{pmatrix}, \begin{pmatrix} \mathbf{h}\\ \mathbf{v} \end{pmatrix} \right)\in C_2(X;L^\vee)\oplus Z_1(X;\mathbb{Z}^{2N}) . \tag*{(104)}

between Gosc\mathcal{G}_{\mathrm{osc}} and C2(X;L∨)⊕Z1(X;Z2N)C_2(X;L^\vee)\oplus Z_1(X;\mathbb{Z}^{2N}). Under this isomorphism, the map σ\sigma acts on Gosc\mathcal{G}_{\mathrm{osc}} as

Posc∋σ(αβhv)⟷∂2(α−ΦT(v−xh)β)−(MhMv)∈C1(X;L∨)(105)\mathcal{P}_{\mathrm{osc}}\ni\sigma \begin{pmatrix} \alpha\\ \beta\\ \mathbf{h}\\ \mathbf{v} \end{pmatrix} \longleftrightarrow \partial_2 \begin{pmatrix} \alpha-\Phi^{T}(\mathbf{v}-x\mathbf{h})\\ \beta \end{pmatrix} - \begin{pmatrix} M\mathbf{h}\\ M\mathbf{v} \end{pmatrix} \in C_1(X;L^\vee) \tag*{(105)}

where M=(ΦTQT)M=\begin{pmatrix}\Phi^{T}\\Q^{T}\end{pmatrix} is the matrix of condensed dyons. Therefore σ\sigma acts via the boundary map on the C2(X;L∨)C_2(X;L^\vee) summand, and by the map of coefficients Z2N→−ML\mathbb{Z}^{2N}\xrightarrow{-M}L on the Z1(X;Z2N)Z_1(X;\mathbb{Z}^{2N}) summand. We can now identify our desired homology as

L=ker⁡ϵABim⁡σ=ker⁡∂1im⁡∂2+Z1(X;L)=H1(X;A).(106)\mathcal{L}=\frac{\ker\epsilon_{AB}}{\operatorname{im}\sigma} =\frac{\ker\partial_1}{\operatorname{im}\partial_2+Z_1(X;L)} =H_1(X;\mathcal{A}). \tag*{(106)}

We therefore recover the fact that the logical group is trivial on the infinite plane, and A⊕AA \oplus A on the torus. The logical operators are given by the closed noncontractible ribbon operators for dyons in L∨\mathcal{L}^{\vee}.

It follows the ground state degeneracy on the torus is ∣A∣|\mathcal{A}|. This is due to the following fact: for a GKP code on a finite number of modes with L\mathcal{L} finite, the dimension of the code space is given by GSD=∣L∣\mathrm{GSD} = \sqrt{|\mathcal{L}|}. This is a standard fact in the theory of lattice GKP codes [9, 30], which consist of only discrete stabilizers; it also holds in our context, where some of the stabilizers are continuous. To apply this argument, we first embed each rotor in a real variable degree of freedom stabilized by Z=e2πix^=1Z = e^{2\pi i\hat{x}} = 1. The key point is that, since all of the continuous stabilizer commute, and each independent continuous stabilizer removes one mode, they can be completely disentangled via Gaussian unitaries without affecting the structure of the logical group. Since L\mathcal{L} is finite, the discrete stabilizers act on the remaining modes as a full-rank lattice (otherwise there would be a continuous family of logical operators). Thus the code on the remaining modes is an ordinary multimode GKP code, which satisfies GSD=∣L∣\mathrm{GSD} = \sqrt{|\mathcal{L}|}.

References

  1. [1]A. Yu. Kitaev, Fault-tolerant quantum computation by anyons, Ann. Phys. 303, 2 (2003).DOI
  2. [2]E. Dennis, A. Kitaev, A. Landahl, and J. Preskill, Topological quantum memory, J. Math. Phys. 43, 4452 (2002).DOI
  3. [3]T. D. Ellison, Y.-A. Chen, A. Dua, W. Shirley, N. Tantivasadakarn, and D. J. Williamson, Topological stabilizer models of twisted quantum doubles, PRX Quantum 3, 010353 (2022).DOI
  4. [4]M. Levin, Protected edge modes without symmetry, Phys. Rev. X 3, 021009 (2013).
  5. [5]M. A. Levin and X.-G. Wen, String-net condensation: A physical mechanism for topological phases, Phys. Rev. B 71, 045110 (2005).
  6. [6]M. Sun, Z. Wang, N. Tantivasadakarn, and Y.-A. Chen, Majorana-Pauli stabilizer codes and duality webs of fermionic topological phases (2026), arXiv:2606.25048.
  7. [7]B. Ruba and B. Yang, Witt groups and bulk-boundary correspondence for stabilizer states (2025), arXiv:2509.10418.arxiv.org/abs/2509.10418
  8. [8]J. C. Magdalena de la Fuente, T. D. Ellison, M. Cheng, and D. J. Williamson, Topological stabilizer models on continuous variables, Phys. Rev. X 16, 011054 (2026).
  9. [9]D. Gottesman, A. Kitaev, and J. Preskill, Encoding a qubit in an oscillator, Phys. Rev. A 64, 012310 (2001).DOI
  10. [10]L. Bhardwaj, D. Gaiotto, and A. Kapustin, State sum constructions of spin-TFTs and string net constructions of fermionic phases of matter, J. High Energy Phys. 04, 096.
  11. [11]F. A. Bais and J. K. Slingerland, Condensate-induced transitions between topologically ordered phases, Phys. Rev. B 79, 045316 (2009).
  12. [12]V. V. Nikulin, Integral symmetric bilinear forms and some of their applications, Math. USSR-Izv. 14, 103 (1980).
  13. [13]A. Davydov, M. Müger, D. Nikshych, and V. Ostrik, The Witt group of non-degenerate braided fusion categories, J. Reine Angew. Math. 677, 135 (2013).arxiv.org/abs/1009.2117
  14. [14]D. Gaiotto, A. Kapustin, N. Seiberg, and B. Willett, Generalized global symmetries, J. High Energy Phys. 02, 172.
  15. [15]G. T. Horowitz, Exactly soluble diffeomorphism invariant theories, Commun. Math. Phys. 125, 417 (1989).
  16. [16]X.-G. Wen and A. Zee, Classification of Abelian quantum Hall states and matrix formulation of topological fluids, Phys. Rev. B 46, 2290 (1992).
  17. [17]D. Belov and G. W. Moore, Classification of abelian spin Chern-Simons theories (2005), arXiv:hep-th/0505235.DOI
  18. [18]X.-G. Wen, Topological orders and edge excitations in fractional quantum Hall states, Adv. Phys. 44, 405 (1995).DOI
  19. [19]In principle, it would suffice to introduce a single mode on each vertex. We introduce a mode for each fermionic generator as a matter of convenience.
  20. [20]S. Seifnashri, Lieb-Schultz-Mattis anomalies as obstructions to gauging (non-on-site) symmetries, SciPost Phys. 16, 098 (2024).
  21. [22]Y.-A. Chen, A. Kapustin, and D. Radicević, Exact bosonization in two spatial dimensions and a new class of lattice gauge theories, Ann. Phys. 393, 234 (2018).
  22. [23]The powers Ãᵖ and B̃ᵠ are defined by multiplying every Pauli exponent in the expressions for à and B̃ by ϕ and q, respectively, before simplifying any factors that equal the identity. For instance, a rotor factor Zᵐ with m ∈ ℤ contributes Zᵐᵠ to Ãᵖ.
  23. [24]M. Qi and W. Shirley, In preparation.
  24. [25]J. Haah, L. Fidkowski, and M. B. Hastings, Nontrivial quantum cellular automata in higher dimensions, Commun. Math. Phys. 398, 469 (2023).
  25. [26]J. Haah, Clifford quantum cellular automata: Trivial group in 2D and Witt group in 3D, J. Math. Phys. 62, 092202 (2021), erratum: J. Math. Phys. 63, 079901 (2022).DOI
  26. [27]W. Shirley, Y.-A. Chen, A. Dua, T. D. Ellison, N. Tantivasadakarn, and D. J. Williamson, Three-dimensional quantum cellular automata from chiral semion surface topological order and beyond, PRX Quantum 3, 030326 (2022).
  27. [28]B. Yang, Categorifying Clifford QCA, Commun. Math. Phys. 407, 77 (2026).arxiv.org/abs/2504.14811
  28. [29]J. Haah, Commuting Pauli Hamiltonians as maps between free modules, Commun. Math. Phys. 324, 351 (2013).
  29. [30]S. Chakraborty and V. V. Albert, Stabilizer codes over general phase spaces (2026), arXiv:2610.04694.arxiv.org/abs/2610.04694

Paper details

vlnc:W_wmasfg9x
Version
v1 · latest
Published
Fresh — added to the Hub within the last 30 days
Added
License
CC BY 4.0
Usage

views · 7

downloads · 0

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

  • Physics and Astronomy

Not recorded for this paper

View statement
Looking for related papers…
Contents