arXiv:2610.15893
Local-dimension-invariant low-density parity-check stabilizer codes
Md. Shahinul Islam, Aaryan Reza, Lane G. Gunderman
Abstract
The protection of quantum information in systems leveraging more than two basis states (qudits) is an ongoing challenge. A tool for transforming the plethora of qubit stabilizer codes into qudit stabilizer codes is the local-dimension-invariant (LDI) extension to the stabilizer formalism. A drawback of the LDI method is that generally in order to ensure the distance of the resultant code is at least preserved the newly selected local-dimension will need to be any prime bigger than a large cutoff value. In this work we consider the repercussions of this large cutoff value, then by focusing on quantum low-density parity-check (qLDPC) codes we are able to bring down this cutoff value. Notably, we are able to reduce it in some cases to the initial local-dimension and for arbitrary CSS qubit-derived codes to singly exponential in $d$ with no dependence on the number of logicals $k$. This reduces the size of the caveat in LDI forms in their distance preservation condition.
AI Summary
Local-dimension-invariant (LDI) stabilizer representations use integer check matrices whose symplectic products vanish exactly, allowing their reduction to commuting checks in any local dimension. The question addressed is how large a new prime dimension must be to guarantee preservation of the source code’s distance, particularly for sparse CSS codes. The paper develops sign-assignment constructions and bounds on check entries, weights and degrees, then relates distance preservation to bounds on determinants of check-matrix minors. For an LDI CSS code of distance $d$, maximum absolute entry $B$, and $\mathcal W$ equal to the smaller of maximum check weight and maximum register degree, the stated sufficient cutoffs include $p^*=B^{d-1}2^{(d-1)(\mathcal W-1)/3}$ for constant sparsity and $p^*\leq(B\mathcal W)^{d-1}$ when $\mathcal W<d$. These use sparse determinant inequalities and row/column-sum bounds. A CSS lifting construction gives $B\leq\lfloor q/2\rfloor^2\min(w_X,w_Z)$. Combined with qubit weight reduction, the paper states a cutoff no larger than $15^{d-1}$, independent of the logical count. Additional constructions concern hypergraph products, sign-only lifts with $B=1$, and Euclidean and hyperbolic surface codes. A Gaussian nearest-boundary model connects increasing GKP dimension to an additional squeezing requirement of $10\log_{10}(p/p_0)$ dB at fixed crossing probability. The guarantees generally remain exponential in distance and sufficient rather than necessary. Sign-assignment conditions are restrictive; the broader cycle condition requires enumeration. The paper also notes Gaussian-elimination fill-in, the lack of a general LDI weight-reduction procedure, and the need for separate distance verification below the cutoff.
Central claim may not hold: The sparsity lemmas assume bounded fill-in under systematicization, although sparse Gaussian elimination can produce column degrees growing with the code length, leaving the arbitrary-CSS cutoff argument unsupported. The first two cutoff formulas apply binary determinant bounds to signed integer LDI matrices, for which those bounds are false even when the maximum entry magnitude is one. The classical systematic lifting prescription can reduce distance from 5 over F_2 to at most 4 over F_3, invalidating the supplied argument for distance preservation in the hypergraph-product construction. The main distance-preservation proof does not control the stabilizer-intersection dimensions, and an integer-commuting CSS lift can turn a source-field low-weight stabilizer into a logical operator over every larger prime.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 14 pages, 1 figure