New local stabilizer codes from local classical codes
Lane G. Gunderman · 2025
Read New local stabilizer codes from local classical codes by Lane G. Gunderman (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-20717.txt.
Introduction
Lane Gunderman builds new quantum error-correcting codes from recent advances in classical two-dimensional codes. His 2025 paper works with stabilizer codes, whose checks watch for errors without reading the stored data. The famous surface code checks locally but stores only one unit of information no matter how large it grows.
Gunderman's codes keep checks local and light, touching only five or six qubits each, while storing more information or guarding it better. One family, tied to the Fibonacci pattern, grows its storage and protection with lattice size. The search behind the work swept translated classical generators across grid sizes up to seventeen by seventeen. The author presents the haul as a fuller map of what local quantum codes can do.
Catalog record
| CREATOR | Lane G. Gunderman |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 20 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2503-20717
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
New local stabilizer codes from local classical codes Lane G. Gunderman Abstract Amongst quantum error-correcting codes the surface code has remained of particular promise as it has local and very low-weight checks, even despite only encoding a single logical qubit no matter the lattice size. In this work we discuss new local and low-weight stabilizer codes which are obtained from the recent progress in 2 D 2D local classical codes. Of note, we construct codes with weight and qubit use count of 5 5 while being able to protect the information with high distance, or greater logical count. We also consider the Fibonacci code family which generates weight and qubit use count of 6 6 while having parameters [ [ O ( l 3 ) , O ( l ) , Ω ( l ) ] ] [[O(l^{3}),O(l),\Omega(l)]] . While other weight-reduction methods centered on lowering the weight without regard to locality, this work achieves very low-weight and geometric locality. This work is exhaustive over translated classical generators of size 3 × 3 3\times 3 and up to size 17 × 17 17\times 17 classical bit grids. Preprint: APS/123-QED I Introduction Quantum mechanical interactions are generated through geometrically local interactions, be it between photons, electron orbitals, or other modes. This naturally motivated means of protecting quantum information using checks that are also geometrically local. Besides this, another natural motivation is to use fewer body interactions as precise control of many-body interactions becomes increasingly challenging as the number of interactions increases. This latter aim led to ideal methods for protecting quantum information requiring few body interactions, which in terms of quantum error-correcting codes corresponds to low-weight checks. These pair of restrictions led to the dominance of the surface code in quantum error-correcting codes as the checks are localized in 2D space using 4 neighboring additional qubit interactions in each check [ 1 ] . While these properties of the code permit it to have a high threshold, and recently experimentally achieve prolonged storage of information, no matter the size of the surface patch used only a single logical qubit is encoded always [ 2 ] . This low logical encoding then requires lattice surgery to perform multi-qubit operations, which can increase the overhead a fair amount [ 3 , 4 ] . There exist non-Euclidean lattices which provide for more encoding, however, their physical implementability for some platforms would be very challenging [ 5 ] . A radically different approach is to drastically increase the encoding rate while retaining the low-weight property of the checks, but permitting the geometric locality to be lost. These result in non-local quantum low-density parity-check (qLDPC) codes 1 1 1 Formally speaking, the surface code is a qLDPC code, but here we mean a non-localized code . While these can have low-weight and higher encoding rate, the fact that the checks are so spread out can make performing these checks challenging. The very best qLDPC codes even require a large amount of far range checks [ 7 , 8 , 9 ] . Given this difficult trade-off, a myriad of different approaches have evolved which aim to keep locality while reducing the weight of the checks–amongst these approaches include hyperboic lattice codes [ 5 ] , subsystem codes [ 10 , 11 , 12 ] , floquet codes [ 13 , 14 ] , and weight reduction methods [ 15 , 16 , 17 ] . Here we restrict ourselves to true stabilizer codes (the quantum analog of classical additive codes), which are believed to be the most promising approach and have a number of nice fault-tolerant features known for them. In this work, we primarily focus on smaller code instances which: 1) are purely geometrically local, 2) have very low weight, 3) encode more than a single logical register. This work leverages the recent results on 2D local classical codes to obtain these results, however, we provide some further analysis of the codes which are possible and illustrate their possible great utility for protecting quantum information [ 18 ] . The work is organized as follows. In the next section definitions are laid out. Following this we show our construction and provide various instances with particularly appealing properties as well as families for which this method applies, making comparisons with other promising results. We then close out by indicating possible future directions to extend this work. II Definitions This work focuses on balancing locality, weight, and the parameters of quantum codes. Since this is achieved through using classical codes, we begin by defining the parameters for classical additive error-correcting codes. The parameters are a triplet of values: n specifies how many classical registers (bits) are used, k the number of logical registers (bits) which are bases for the null space of the parity check matrix, and d the distance which is the minimal number of columns which are linearly dependent. These are written as [n,k,d] . In the quantum case we assume that the given codes have been written in the symplectic representation so that the codes are represented as a matrix of powers of the Pauli operators [ 19 ] . Using this representation, the parameters for the quantum code are rather similar to the classical code: n is the number of physical registers (qubits, or more general), k is the number of logical registers which can be deduced from the number of physical registers minus the count of the symplectically independent rows of the check matrix, and d is the distance of the code which is the smallest number of columns that are symplectically linearly dependent, but does not correspond to one of the checks. These are written as [[n,k,d]] . A final element needed for the results in this work is our selected method to transform classical codes into quantum codes. Let H_{1} be a classical linear code with r_{1} rows and parameters [n_{1},k_{1},d_{1}] and let H_{2} be a classical linear code with r_{2} rows and parameters [n_{2},k_{2},d_{2}] . Then we may form the hypergraph product stabilizer code from these codes with parity checks given by H_{X}=(H_{1}\otimes I_{n_{2}}\quad I_{r_{1}}\otimes H_{2}^{T}) and H_{Z}=(I_{n_{1}}\otimes H_{2}\quad H_{1}^{T}\otimes I_{r_{2}}) 2 2 2 For the qudit version, or generally local-dimension-invariant form, we simply put a minus sign in front of one term so that the symplectic product is perfectly zero. [ 21 ] . The resulting code will have parameters [[n_{1}n_{2}+r_{1}r_{2},k_{1}k_{2}+k_{1}^{T}k_{2}^{T},\min(d_{1},d_{2},d_{1}^{T},d_{2}^{T})]] . II.1 Notation Throughout this work we use alphabetically ordered strings of the letters a to i to indicate the cyclically generating local 2D classical checks where the letters indicate support within the following 3\times 3 fundamental check block: \begin{pmatrix}a&b&c\\ d&e&f\\ g&h&i\end{pmatrix}. (1) As a simple example the string "ab" would indicate a set of repetition codes whereby adjacent registers are used in the check, then "ab" is translated across the 2D set of registers to generate the full set of checks. When discussing particular instances we augment this notation for generators with a subscript indicating the dimension of the grid of registers in the form w\times h (width by height), and following the string with the traditional classical error-correcting code parameter notation [n,k,d] . In this work we will only focus