Description
Galois-qudit code that utilizes the notion of a lifted product in its construction. Lifted products of certain classical Tanner codes are the first (asymptotically) good QLDPC codes.
A code can be defined by \(LP(A,B)\), where \(A\) and \(B\) are a pair of matrices with elements from a group algebra. Heuristically, the code is a hypergraph product over the group algebra \(\mathbb{F}_q[G]\), with each entry then replaced by a \(|G|\times|G|\) matrix. The code thus has \(|G|\) times as many qudits and checks as a hypergraph product of matrices of the same shapes [2; Sec. III]. For \(q=2\), each block’s row and column weight equals the number of group elements in its entry [2; Sec. III]. Entries that are single group elements therefore keep the check weights of a hypergraph product with ones in their place.
More technically, a lifted product over a ring \(R\) is a product of two chain complexes whose chains are free modules over \(R\). An interesting case is when \(R=\mathbb{F}_q [G]\), the group-\(G\) algebra over the finite field \(\mathbb{F}_q\); in this case, the product can be called a \(G\)-lifted product. Just like its further generalization the balanced product, a lifted product code generalizes a hypergraph product code in that a reduction of symmetry is exploited to decrease the number of physical qubits required. The first version of this construction contains hypergraph-product codes in the case where one of the two input parity-check matrices is square [1].
The key operation behind the \(G\)-lifted product is the \(G\)-lift, a group-algebraic version of the lifting procedure of protograph LDPC codes. The lifted product is obtained from the usual hypergraph product of the two lifted complexes by quotienting by the shared \(G\) action [3].
Lifted products over non-Abelian group algebras have been studied, e.g., over the dihedral group algebra \(\mathbb{F}_q[D_{2n}]\), whose Wedderburn decomposition yields short non-Abelian quantum moderate-density parity-check codes—dubbed dihedral quantum codes—with an explicit dimension formula and distance bound [4]. Non-Abelian lifts also underlie mitten codes, whose non-commutativity removes a distance ceiling that constrains Abelian LP codes of the same base-matrix shape [5].
Protection
Code performance strongly depends on the group \(G\) used in the product [2]. For \(q=2\), the distance of \(LP(A,B)\) is at most the smaller distance of its two classical base codes [5; Thm. 8]. This requires wide base matrices \(A\) and \(B\) whose left and right regular representations, respectively, have full row rank. A \([[200,13,10]]\) LP code over \(\mathbb{F}_2[C_4\times C_2]\) whose base codes have distance eight shows that the full-rank condition cannot be dropped [5; Remark 12]. For a one-row base matrix, the base-code distance is at most \((\mathrm{wt}(a_i)+\mathrm{wt}(a_j))\,|[G,G]|\) for any two entries \(a_i,a_j\) [5; Prop. 4]. The bound requires that the sum of the elements of the commutator subgroup \([G,G]\) not annihilate both entries. An entry \(1+g\) in a one-row base matrix bounds the base-code distance, and hence the LP-code distance, by the order of \(g\) [5; Thm. 9].Decoding
Linear time iterative decoder [6].Cousins
- Haah cubic code (CC)— A lifted-product code constructed with coefficients in the ring \(R=\mathbb{F}_2[x,y,z]/(x^L-1,y^L-1,z^L-1)\) is a cubic code [3; Appx. B].
- Actively orthogonal CSS code— LP codes enforce the CSS orthogonality condition as an algebraic identity on all rows, whereas actively orthogonal CSS codes impose it only on the retained rows [7,8].
- Fiber-bundle code— The specific fiber-bundle QLDPC code achieving a distance scaling better than \(\sqrt{n}~
polylog (n)\) can also be formulated directly as an LP code (see published version of Ref. [9]). Lifted products of a length-one with a length-\(m\) chain complex can be thought of as fiber-bundle codes [10]. - Toric code— A lifted-product code for the ring \(R=\mathbb{F}_2[x,y]/(x^L-1,y^L-1)\) is the toric code [3; Appx. B].
- Subsystem lifted-product (SLP) code— SLP codes reduce to (subspace) LP codes when there is no gauge subsystem.
- Two-block CSS code— LP codes can be constructed using non-square matrices and taking a hypergraph product over a group algebra, while two-block CSS codes are constructed directly using square matrices.
Primary Hierarchy
References
- [1]
P. Panteleev and G. Kalachev, “Degenerate Quantum LDPC Codes With Good Finite Length Performance”, Quantum 5, 585 (2021) arXiv:1904.02703 DOI
- [2]
P. Panteleev and G. Kalachev, “Quantum LDPC Codes With Almost Linear Minimum Distance”, IEEE Transactions on Information Theory 68, 213 (2022) arXiv:2012.04068 DOI
- [3]
P. Panteleev and G. Kalachev, “Asymptotically Good Quantum and Locally Testable Classical LDPC Codes”, (2022) arXiv:2111.03654
- [4]
N. Willenborg, M. Borello, A.-L. Horlemann, and H. Islam, “Dihedral Quantum Codes”, (2025) arXiv:2310.15092
- [5]
A. Bhardwaj, M. Ma, N. Meister, R. King, D. Bluvstein, J. Preskill, M. Cain, Q. Xu, and H.-Y. Huang, “High-rate qLDPC processors”, (2026) arXiv:2607.28795
- [6]
A. K. Pradhan, N. Raveendran, N. Rengaswamy, and B. Vasić, “Linear Time Iterative Decoders for Hypergraph-Product and Lifted-Product Codes”, (2025) arXiv:2504.01728
- [7]
K. Kasai, “Breaking the Orthogonality Barrier in Quantum LDPC Codes”, Quantum 10, 2205 (2026) arXiv:2601.08824 DOI
- [8]
W. Yang, C. Duckering, and A. Dua, “Designer Codes from GALA: Compact, Self-Dual, and Rate-1/2 QEC on Reconfigurable Atom Arrays”, (2026) arXiv:2608.07431
- [9]
M. B. Hastings, J. Haah, and R. O’Donnell, “Fiber bundle codes: breaking the n \({}^{
1/2 }\) polylog( n ) barrier for Quantum LDPC codes”, Proceedings of the 53rd Annual ACM SIGACT Symposium on Theory of Computing 1276 (2021) arXiv:2009.03921 DOI - [10]
N. P. Breuckmann and J. N. Eberhardt, “Balanced Product Quantum Codes”, IEEE Transactions on Information Theory 67, 6653 (2021) arXiv:2012.09271 DOI
- [11]
Y. Tan, B. Roberts, N. Tantivasadakarn, B. Yoshida, and N. Y. Yao, “Fracton models from product codes”, Physical Review Research 7, (2025) arXiv:2312.08462 DOI
- [12]
P. Panteleev and G. Kalachev, “Maximally Extendable Sheaf Codes”, (2024) arXiv:2403.03651
Page edit log
- Victor V. Albert (2026-11-03) — most recent
- Victor V. Albert (2026-09-30)
- Victor V. Albert (2026-09-26)
- Victor V. Albert (2026-08-24)
- Victor V. Albert (2026-08-17)
- Victor V. Albert (2026-08-15)
- Victor V. Albert (2026-07-18)
- Victor V. Albert (2026-06-08)
- Victor V. Albert (2022-01-17)
- Finnegan Voichick (2021-12-14)
- Pavel Panteleev (2021-11-30)
Cite as:
“Lifted-product (LP) code”, The Error Correction Zoo (V. V. Albert & P. Faist, eds.), 2026. https://errorcorrectionzoo.org/c/lifted_product, arXiv:2606.11484