Logically Augmented Quantum LDPC Codes: Construction and Hierarchical Decoding

Ayushi Shirke; Hessam Mahdavifar; Ivana Dimitrova; Sheida Rabeti

REVIEW

Logically Augmented Quantum LDPC Codes: Construction and Hierarchical Decoding

Sheida Rabeti, Ayushi Shirke, Ivana Dimitrova, Hessam Mahdavifar

Information Theory · Submitted 4 Nov 2026 · 31 references

Pith's verdict: A clean in-block expurgation construction with a genuinely neat two-hypothesis repetition trap, carrying an empirical headline that is not trustworthy until the logical-operator weights and baselines are reported.

T0 · AI review

How this review was made: An AI referee (deepseek/deepseek-v4-flash) read the paper on November 6, 2026 and answered a fixed list of questions about it. T0 is the starting level every paper gets. The levels describe how deeply Pith has checked the paper, never the authors.

T0 AI review An AI read the paper and answered a fixed list of questions.
T1 Audit Automatic checks that the references and files the paper relies on exist and match.
T2 Compute The paper's code or calculations re-run with the same results.
T3 Formal statement The claims written out in a formal math language a computer can read.
T4 Proof Those claims proved, and the proof checked by computer.

How Pith checks papers →

· 2 major objections · 3 minor · Ask about this paper · The compiled record · Read the full review →

Challenge this review Comment (0) Read on arXiv ↗

Cite ▾

arXiv ID2611.07062 (version 1)

Pith ID4FLQINUK, Pith's permanent record of this paper and its review

CitationAyushi Shirke, Hessam Mahdavifar, Ivana Dimitrova, Sheida Rabeti. "Logically Augmented Quantum LDPC Codes: Construction and Hierarchical Decoding." arXiv:2611.07062 (2026). https://pith.science/paper/4FLQINUK

Cite this reviewBibTeX for Pith's review of this paper

Claim paper

Why Pith says this

Highlight any text to comment on it.

Editor's note

read the letter

The construction is the real contribution; the empirical headline is the soft spot.

What's actually new: Definition 1—LPC-structured expurgation inside a single block, preserving block length and every original stabilizer generator—and Example 2, the observation that a repetition LPC admits exactly two logical error hypotheses regardless of length, capping the exhaustive list at two conditional BP runs. That's a clean, workable rate-versus-protection knob for existing QLDPC codes, and the algebra is sound: the orthogonality conditions are sufficient for valid CSS augmentation, the parameter arithmetic checks out ([[2610,744]] with r=50 gives [[2610,695]], 6.59% rate cost), and nothing is fitted to the claims. The paper is honest about scope—three code families, biased noise explicitly left to future work—and I don't hold that against it.

The reader's CONDITIONAL verdict is about right, and the stress-test lands. The load-bearing claim is the Section V headline: up to 81/75/69% LER reduction from HLLD over direct BP of the same LA code. The trouble is Remark 1 only upper-bounds the added check weights by 2w_Z and never reports w_Z, the L_Z representatives, or the selection procedure. If those logical operators aren't light, the augmented matrix stops being LDPC, and direct BP runs on checks of weight up to 2w_Z while HLLD conditions on single L_Z rows. Sum-product BP degrades with check degree, so the reported gains may reflect a handicapped baseline rather than the benefit of list conditioning. That is a reproducibility gap in the central quantitative claim, and it is fixable: report w_Z and the selection, and add a baseline robust to check weight.

I'd downgrade one of the reader's four asks. The missing d_A proof is minor—expurgation only shrinks the kernel, so d_A is at least the primary distance, and the dimension count is routine from the stated conditions. The other two stands: error bars would help (100 failures is a thin floor), and there is no code or data release, so the headline numbers cannot be independently checked.

Who gets value: anyone working on QLDPC decoding or biased-noise memory engineering benefits from the construction and the two-hypothesis trick. The paper deserves a serious referee. The revision ask is data and baselines, not new theory—send it out, and expect a revision.

Referee report

Significance

The algebraic core is sound and clean: the stated conditions in Definition 1 are sufficient for CSS orthogonality, and the dimension and rate numbers in Table I and Section V check out, e.g., r=50 on [[2610,744]] gives [[2610,695]], a 6.59% rate reduction. The HLLD mechanism is a reasonable and low-overhead way to exploit the LPC hierarchy: for repetition LPCs the logical list has exactly two hypotheses, so the added decoding cost is at most two conditioned BP runs, and the numerical comparisons are measured rather than fitted. The main value is the joint construction-decoding perspective, which allows the same primary code to be adapted to a rate-protection target without changing the block length. The strength of the quantitative claims, however, is currently limited by missing information about the weights and selection of the logical representatives used in the augmented stabilizers; this determines whether the examples are genuinely LDPC and whether the direct-BP baseline is a fair comparator.

Major objections (2)

[Section V, Figs. 2-4; Remark 1] The headline 69-81% improvements over direct BP cannot be fully assessed from the data as presented, because the paper never reports the actual weights of the selected logical representatives L_Z, nor the procedure used to select them. Remark 1 only provides the upper bound w(H_A^Z) <= max{w(H_Z), 2w_Z}, and this is an upper bound, not data. The direct-BP baseline runs sum-product decoding on the augmented matrix whose added Z-checks have weight up to 2w_Z, while HLLD's conditioned matrix H_C in Eq. (4) uses the L_Z rows themselves, of weight w_Z. If w_Z is not small, the LA code is no longer LDPC and direct BP is expected to degrade from high check degree alone, so the reported gains may reflect a handicapped baseline rather than the benefit of hierarchical list conditioning. Please report w_Z, the selection method for the logical representatives, and the resulting actual augmented check weights for the lifted-product, bivariate-bicycle, and mitten examples.
[Definition 1, Section III] Definition 1 states that the LA code has parameters [[n, k-n_L+k_L, d_A]], but the manuscript does not prove that the added rows G_X L_X and G_Z L_Z are independent of the original stabilizer rows or that d_A is at least the primary-code distance. The dimension identity is load-bearing because it is used for every code parameter quoted in Section V, including [[2610,695]], [[144,10]], and [[780,147]]. The argument is short: under full row rank of G_X and G_Z, the selected logical representatives are independent modulo the original row spaces, and expurgation cannot decrease either CSS distance. Please include this verification explicitly in Section III, and state the lower bound d_A >= d so the notation d_A is informative.

Minor objections (3)

[Section V] The simulation section states a stopping rule of at least 100 decoding failures but does not report the total number of Monte Carlo trials or confidence intervals; adding this information would make the claimed reductions, e.g., the 81% figure in Fig. 4, easier to assess.
[Section IV-A, Eq. (4)] It would help to state explicitly that H_C is a decoder-side conditioning matrix rather than a stabilizer parity-check matrix of the LA code, since the L_Z rows are logical operators used as pseudo-checks during conditioned BP.
[Section IV-C] The complexity expression O(I_max |E_P| + T_ListGen + L I_max |E_C|) omits the cost of OSD when HLLD is combined with BP-OSD; the OSD post-processing cost inside the conditioned branches should be included or stated as a separate term.

How Pith checks papers → · Full audit of this paper →

The paper

The paper claims

Layer a small auxiliary code onto selected logical qubits of an existing quantum LDPC code, and a hierarchical list decoder can cut the logical error rate by up to 81 percent without adding physical qubits.

The reading

The paper argues that a quantum LDPC code need not be redesigned to change its rate or its balance of error protection: an off-the-shelf CSS code can be augmented by imposing additional stabilizer constraints on a few of its logical qubits, using a small auxiliary Logical Protection Code (LPC), to obtain a new code $[[n, k - n_L + k_L, d_A]]$ with the same physical block length and the same original stabilizer generators. With a repetition code as the LPC, each added stabilizer is a product of just two logical operators and the rate loss is a tunable knob. The second claim is that this structure should be exploited by the decoder: when belief propagation on the primary code fails to converge, the proposed Hierarchical Logical List Decoder (HLLD) uses the LPC syndrome to enumerate logical error hypotheses, runs a conditioned BP decode under each, and returns the lightest syndrome-consistent correction. Against direct BP on the same augmented code, the simulations report logical-error-rate reductions of up to 81%, 75%, and 69% for mitten-, lifted-product-, and bivariate-bicycle-based LA codes, and a 2.5x improvement at a 5.24% rate cost for repetition length $r = 40$.

What carries the argument

The load-bearing object is the logical error vector $u = L_Z e^T$, the image of the physical error under the selected logical $Z$ operators, together with the measurable LPC syndrome $t = G_Z u$, which pins $u$ down only up to the LPC's kernel cosets. The decoder's workhorse is the conditioned parity-check matrix $H_C = [H_Z; L_Z]$ with syndrome $[s; u^{(\ell)}]$, which forces belief propagation to search under one specific logical error hypothesis; for a repetition LPC the hypothesis space has exactly two elements. On the construction side, the repetition LPC's check rows each have weight two, so every added stabilizer is the product of two selected logical operators and the augmented matrix's row weight is bounded by $\max\{w(H_Z), 2w_Z\}$ as stated in Remark 1.

Extended reading notes

Core claim

The central discovery, on the paper's own terms, is that logical-level redundancy can be injected into an existing CSS QLDPC code without touching the physical layer: given a primary code $[[n, k, d]]$ and $n_L$ selected pairs of independent logical operators with physical representatives $L_X, L_Z$, any CSS LPC $[[n_L, k_L, d_L]]$ with parity-check matrices $G_X, G_Z$ yields an augmented code whose parity-check matrices are $[H_X; G_X L_X]$ and $[H_Z; G_Z L_Z]$. All original stabilizer generators and the block length $n$ are preserved, so the only cost of the upgrade is the rate reduction $(n_L - k_L)/n$. A repetition LPC of length $r$ gives $[[n, k - r + 1, d_A]]$, and because the LPC's parity-check matrix has rank $r - 1$, every LPC syndrome admits exactly two logical error hypotheses regardless of $r$; HLLD conditions a BP run on each and keeps the lighter syndrome-consistent correction. The reported consequences are a tunable rate-protection trade-off, with a factor-2.5 logical-error-rate improvement at a 5.24% rate reduction for the [[2610,744]] lifted-product code at $r = 40$, and decoding gains of 81%, 75%, and 69% over direct BP of the same LA codes built from mitten, lifted-product, and bivariate bicycle primary codes.

Load-bearing premise

The quantitative gains presuppose that each primary code supplies logical $Z$ and $X$ operators light enough that the added stabilizer rows stay sparse: Remark 1 bounds the augmented row weight by $\max\{w(H_Z), 2w_Z\}$ but the paper never reports the actual weights or the selection procedure for $L_Z$, and heavy logical operators would make the augmented code non-LDPC while handicapping the direct-BP decoder used as the baseline.

Editorial extensions

If this is right

Any existing QLDPC primary code can be upgraded in place: the block length and every original stabilizer generator survive the augmentation, so the LA code inherits the primary code's hardware footprint and syndrome-extraction circuitry.
Repetition length is a rate-protection dial: longer $r$ buys a better logical error rate at a linearly growing rate cost, while the decoder's list size stays at two because each repetition syndrome has exactly two preimages.
The LPC plays a double role, supplying both the extra stabilizer constraints and the logical hypotheses that organize the decoding search, so the hierarchy is part of the code's definition rather than an add-on.
The reported decoding gains hold across three structurally different code families (mitten, lifted-product, bivariate bicycle) and extend into the low-error regime when BP-OSD serves as the constituent decoder.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

If short logical representatives exist for the target families, the construction is a generic post-hoc rate-shaping layer: a code already running on hardware could be converted into a lower-rate, higher-protection version, and choosing an X- versus Z-oriented LPC could bias protection toward the dominant physical error type, a knob the paper leaves to future work.
The two-hypothesis property suggests a scaling test the paper does not run: compare HLLD-FULL against a decoder that folds the two logical hypotheses into a single translated decoding problem on one Tanner graph, to separate the gain of hypothesis diversity from the gain of conditioned BP.
Replacing the repetition LPC with a short higher-rate LPC would grow the exhaustive logical list from two to $2^{n_L - \mathrm{rank}(G_Z)}$ candidates; the paper's framework allows this but gives no numerics for it, so the point at which exhaustive lists stop paying for their conditioned BP branches is an open question.

Original abstract

We propose a joint framework for constructing and decoding Logically Augmented (LA) quantum low-density parity-check (QLDPC) codes. This is motivated by the need for flexible coding rates and adaptive error protection while preserving block length and code structure. On the construction side, we use an auxiliary \textit{Logical Protection Code (LPC)} to impose additional stabilizer constraints on selected logical qubits of a \textit{primary} code. On the decoding side, the proposed decoder uses the LPC syndrome to generate logical error hypotheses and condition belief propagation (BP) on each hypothesis. The proposed \textit{Hierarchical Logical List Decoder (HLLD)} is invoked only when the initial BP decoder for primary code fails to converge. It then selects the correction with the smallest Hamming weight among the syndrome-consistent candidates produced by hypothesis-conditioned BP runs. For numerical evaluation, we focus on short repetition LPCs, which provide a flexible means of trading rate for additional error protection. Each LPC syndrome admits exactly two logical error hypotheses, irrespective of repetition length, which enables HLLD to explore both possibilities without relying on a single logical estimate. Numerical results show that varying the repetition length provides a tunable trade-off between rate and error-correction performance across a wide range of primary code constructions and their operating regimes. On the decoding side, with the overall LA code held fixed, HLLD reduces the logical error rate by up to 81\%, 75\%, and 69\% for LA codes based on mitten, lifted-product, and bivariate bicycle codes, respectively, compared with conventional BP decoding of the same LA codes.

The compiled record

The paper's own sentences, arranged, kept beside the review and separate from it. Each label is drawn from a closed vocabulary; the number beside it is the probability the classifier gave that label. Numbers in the tables are the authors', checked by code. No sentence in this record was written by a model.

Compiled by classifier-1.13.0 via TypeSafe API, question bank abstract-p1-20260923, 2026-11-06: 1 request, 7,710 input tokens.

What the abstract itself states, as probabilities: claims first 0.04 · states limitation 0.29 · human participants 0.01 · controlled comparison 0.94 · randomized assignment 0.01 · states quantitative result 0.99 · compares with prior results 0.95 · mentions code or data availability 0.02.

The abstract, sentence by sentence

arXiv listing abstract v1. Results and the main claim are underlined; show every role.

We propose a joint framework for constructing and decoding Logically Augmented (LA) quantum low-density parity-check (QLDPC) codes. This is motivated by the need for flexible coding rates and adaptive error protection while preserving block length and code structure. On the construction side, we use an auxiliary \textit{Logical Protection Code (LPC)} to impose additional stabilizer constraints on selected logical qubits of a \textit{primary} code. On the decoding side, the proposed decoder uses the LPC syndrome to generate logical error hypotheses and condition belief propagation (BP) on each hypothesis. The proposed \textit{Hierarchical Logical List Decoder (HLLD)} is invoked only when the initial BP decoder for primary code fails to converge. It then selects the correction with the smallest Hamming weight among the syndrome-consistent candidates produced by hypothesis-conditioned BP runs. For numerical evaluation, we focus on short repetition LPCs, which provide a flexible means of trading rate for additional error protection. Each LPC syndrome admits exactly two logical error hypotheses, irrespective of repetition length, which enables HLLD to explore both possibilities without relying on a single logical estimate. Numerical results show that varying the repetition length provides a tunable trade-off between rate and error-correction performance across a wide range of primary code constructions and their operating regimes. On the decoding side, with the overall LA code held fixed, HLLD reduces the logical error rate by up to 81\%, 75\%, and 69\% for LA codes based on mitten, lifted-product, and bivariate bicycle codes, respectively, compared with conventional BP decoding of the same LA codes.

background contribution method result interpretation main claim

What the classifier reports for each sentence. The number is the probability it gave the label.
#	sentence	role	the authors' own work	states the main conclusion	strength
s0	We propose a joint framework for constructing and decoding Logically Augmented (LA) quantum low-density parity-check (QLDPC) codes.	contribution 1.00	0.90	0.08	asserted 0.98
s1	This is motivated by the need for flexible coding rates and adaptive error protection while preserving block length and code structure.	background 0.86	0.10	0.03	asserted 0.97
s2	On the construction side, we use an auxiliary \textit{Logical Protection Code (LPC)} to impose additional stabilizer constraints on selected logical qubits of a \textit{primary} code.	method 0.94	0.89	0.04	asserted 1.00
s3	On the decoding side, the proposed decoder uses the LPC syndrome to generate logical error hypotheses and condition belief propagation (BP) on each hypothesis.	method 0.97	0.89	0.04	asserted 1.00
s4	The proposed \textit{Hierarchical Logical List Decoder (HLLD)} is invoked only when the initial BP decoder for primary code fails to converge.	method 1.00	0.80	0.04	asserted 1.00
s5	It then selects the correction with the smallest Hamming weight among the syndrome-consistent candidates produced by hypothesis-conditioned BP runs.	method 1.00	0.82	0.06	asserted 1.00
s6	For numerical evaluation, we focus on short repetition LPCs, which provide a flexible means of trading rate for additional error protection.	method 0.92	0.75	0.04	asserted 0.93
s7	Each LPC syndrome admits exactly two logical error hypotheses, irrespective of repetition length, which enables HLLD to explore both possibilities without relying on a single logical estimate.	method 0.93	0.76	0.06	asserted 1.00
s8	Numerical results show that varying the repetition length provides a tunable trade-off between rate and error-correction performance across a wide range of primary code constructions and their operating regimes.	result 0.76	0.95	0.28	asserted 1.00
s9	On the decoding side, with the overall LA code held fixed, HLLD reduces the logical error rate by up to 81\%, 75\%, and 69\% for LA codes based on mitten, lifted-product, and bivariate bicycle codes, respectively, compared with conventional BP decoding of the same LA codes.	result 1.00	0.96	0.55	asserted 1.00

What kind of paper the classifier reports this is: method or tool 1.00 · evidence from simulation or computation 1.00.

What was examined, and what was not

sources read

arXiv listing abstract v1

not examined

full text
figures and tables
references and citing sentences
relations to other papers

levels run

P abstract

not run

S sections
T body sentences
C citations
R cross-paper relations

Every quoted sentence on this record is the authors'. Every label is from Pith's closed vocabulary, with the probability shown beside it. Every number in the measurements table is the authors', checked by code. Record as JSON

Figures

**Figure 1.** Logical-level encoding of the LA CSS code using the

**Figure 2.** Decoding performance of the logically augmented

**Figure 3.** Decoding performance of the logically augmented code constructed from the

**Figure 4.** Decoding performance of the logically augmented

Figures reproduced from arXiv:2611.07062 by the authors.

References

Works cited (31)

[1]

Z. Babar, P. Botsinis, D. Alanis, S. X. Ng, and L. Hanzo, “Fifteen years of quantum LDPC coding and improved decoding strategies,”IEEE Access, vol. 3, pp. 2492–2519, 2015

View on Pith 2015
[2]

N. P. Breuckmann and J. N. Eberhardt, “Quantum low-density parity- check codes,”PRX Quantum, vol. 2, no. 4, Oct. 2021

View on Pith 2021
[3]

B. Vasic, V . Savin, M. Pacenti, S. Borah, and N. Raveendran, “Quantum low-density parity-check codes,” 2025. [Online]. Available: https://arxiv.org/abs/2510.14090

View on Pith arXiv 2025
[4]

S. J. Evered, D. Bluvstein, M. Kalinowski, S. Ebadi, T. Manovitz, H. Zhou, S. H. Li, A. A. Geim, T. T. Wang, N. Maskaraet al., “High- fidelity parallel entangling gates on a neutral-atom quantum computer,” Nature, vol. 622, no. 7982, pp. 268–272, 2023

View on Pith 2023
[5]

S. J. Evered, M. Xu, S. H. Li, A. A. Geim, J. P. Bonilla Ataides, M. Kalinowski, D. Bluvstein, N. Maskara, C. Kokail, M. Greineret al., “High-fidelity entangling gates and nonlocal circuits with neutral atoms,” Science, vol. 394, no. 6820, pp. 197–201, 2026

View on Pith 2026
[6]

E. Knill and R. Laflamme, “Concatenated quantum codes,”arXiv preprint quant-ph/9608012, 1996

View on Pith Pith review arXiv 1996
[7]

M. Grassl, P. Shor, G. Smith, J. Smolin, and B. Zeng, “Generalized concatenated quantum codes,”Physical Review A—Atomic, Molecular, and Optical Physics, vol. 79, no. 5, p. 050306, 2009. 0.01 0.015 0.02 0.03 0.05 0.07 0.10 10−4 10−3 10−2 10−1 100 Physical Error Ratep x Logical Error Rate (a) BP BP, BB [[144, 12]] BP, LA-BB [[144, 10]] HLLD-MAP, LA-BB [[144...

View on Pith 2009
[8]

M. J. Gullans, S. Krastanov, D. A. Huse, L. Jiang, and S. T. Flammia, “Quantum coding with low-depth random circuits,”Physical Review X, vol. 11, no. 3, p. 031066, 2021

View on Pith 2021
[9]

C. Gidney, M. Newman, P. Brooks, and C. Jones, “Yoked surface codes,” Nature Communications, vol. 16, no. 1, p. 4498, 2025

View on Pith 2025
[10]

P. Shanahan and D. Ruiz, “Elevator codes: Concatenation for resource-efficient quantum memory under biased noise,”arXiv preprint arXiv:2601.10786, 2026

View on Pith Pith review arXiv 2026
[11]

P. Panteleev and G. Kalachev, “Degenerate quantum LDPC codes with good finite length performance,”Quantum, vol. 5, p. 585, 2021

View on Pith 2021
[12]

H. Yao, W. A. Laban, C. Häger, A. G. i Amat, and H. D. Pfister, “Belief propagation decoding of quantum LDPC codes with guided decimation,” in2024 IEEE International Symposium on Information Theory (ISIT). IEEE, 2024, pp. 2478–2483

View on Pith 2024
[13]

J. Du Crest, M. Mhalla, and V . Savin, “Stabilizer inactivation for message- passing decoding of quantum LDPC codes,” in2022 IEEE Information Theory Workshop (ITW). IEEE, 2022, pp. 488–493

View on Pith 2022
[14]

S. Koutsioumpas, H. Sayginel, M. Webster, and D. E. Browne, “Auto- morphism ensemble decoding of quantum LDPC codes,”arXiv preprint arXiv:2503.01738, 2025

View on Pith Pith review arXiv 2025
[15]

S. Rabeti and H. Mahdavifar, “Multiple-bases belief propagation list decoding for quantum LDPC codes,” 2026, accepted to IEEE Information Theory Workshop (ITW 2026), available at arXiv:2605.14170

View on Pith Pith review arXiv 2026
[16]

D. Poulin, “Optimal and efficient decoding of concatenated quantum block codes,”Physical Review A—Atomic, Molecular, and Optical Physics, vol. 74, no. 5, p. 052333, 2006

View on Pith 2006
[17]

C. Zhang, Z. Wu, J. Wu, and S. Huang, “Bidirectional decoding for con- catenated quantum hamming codes,”arXiv preprint arXiv:2601.09131, 2026

View on Pith arXiv 2026
[18]

T. Kakizaki, “Approximate level-by-level maximum-likelihood decoding based on the chase algorithm for high-rate concatenated stabilizer codes,” arXiv preprint arXiv:2601.18743, 2026

View on Pith arXiv 2026
[19]

S. Kudekar, T. Richardson, and A. Iyengar, “The effect of saturation on belief propagation decoding of LDPC codes,” in2014 IEEE International Symposium on Information Theory. IEEE, 2014, pp. 2604–2608

View on Pith 2014
[20]

F. J. MacWilliams and N. J. A. Sloane,The theory of error-correcting codes. Elsevier, 1977, vol. 16

View on Pith 1977
[21]

J. Roffe, “LDPC: Python tools for low density parity check codes,”

View on Pith
[22]

J. Roffe, D. R. White, S. Burton, and E. Campbell, “Decoding across the quantum low-density parity-check code landscape,”Physical Review Research, vol. 2, no. 4, Dec. 2020

View on Pith 2020
[23]

M. Cain, Q. Xu, R. King, L. R. Picard, H. Levine, M. Endres, J. Preskill, H.-Y . Huang, and D. Bluvstein, “Shor’s algorithm is possible with as few as 10,000 reconfigurable atomic qubits,”arXiv preprint arXiv:2603.28627, 2026

View on Pith Pith review arXiv 2026
[24]

using a repetition LPC of lengthr= 3: (a) BP and (b) BP-OSD, comparing direct decoding with the proposed HLLD-MAP and HLLD-FULL strategies. 0.02 0.025 0.03 0.04 0.05 10−4 10−3 10−2 10−1 Physical Error Ratep x Logical Error Rate BP, Mitten [[780, 156]] BP, LA-Mitten [[780, 147]] HLLD-MAP, LA-Mitten [[780, 147]] HLLD-FULL, LA-Mitten [[780, 147]] Fig. 4: Dec...

View on Pith
[25]

using a repetition LPC of lengthr= 10, under direct BP and the proposed HLLD-MAP and HLLD-FULL

View on Pith
[26]

A. A. Kovalev and L. P. Pryadko, “Quantum kronecker sum-product low- density parity-check codes with finite rate,”Phys. Rev. A, vol. 88, p. 012311, Jul 2013

View on Pith 2013
[27]

S. Bravyi, A. W. Cross, J. M. Gambetta, D. Maslov, P. Rall, and T. J. Yoder, “High-threshold and low-overhead fault-tolerant quantum memory,”Nature, vol. 627, no. 8005, p. 778–782, Mar. 2024

View on Pith 2024
[28]

A. Bhardwaj, M. Ma, N. Meister, R. King, D. Bluvstein, J. Preskill, M. Cain, Q. Xu, and H.-Y . Huang, “High-rate qLDPC processors,”arXiv preprint arXiv:2607.28795, 2026

View on Pith Pith review arXiv 2026
[30]

R. Wang and L. P. Pryadko, “Distance bounds for generalized bicycle codes,” 2022. [Online]. Available: https://arxiv.org/abs/2203.17216

View on Pith Pith review arXiv 2022
[31]

S. Rabeti and H. Mahdavifar, “Univariate bicycle quantum LDPC codes: Explicit logical structure and distance bounds,” 2026, accepted to IEEE Information Theory Workshop (ITW 2026), available at arXiv:2605.14173

View on Pith Pith review arXiv 2026
[2022]

[Online]. Available: https://pypi.org/project/ldpc/

View on Pith

Discussion

Disagree with the review, or know something it missed? Leave a comment, highlight any passage above to comment on it, or challenge this review formally.

[1] [1]

Z. Babar, P. Botsinis, D. Alanis, S. X. Ng, and L. Hanzo, “Fifteen years of quantum LDPC coding and improved decoding strategies,”IEEE Access, vol. 3, pp. 2492–2519, 2015

View on Pith 2015

[2] [2]

N. P. Breuckmann and J. N. Eberhardt, “Quantum low-density parity- check codes,”PRX Quantum, vol. 2, no. 4, Oct. 2021

View on Pith 2021

[3] [3]

B. Vasic, V . Savin, M. Pacenti, S. Borah, and N. Raveendran, “Quantum low-density parity-check codes,” 2025. [Online]. Available: https://arxiv.org/abs/2510.14090

View on Pith arXiv 2025

[4] [4]

S. J. Evered, D. Bluvstein, M. Kalinowski, S. Ebadi, T. Manovitz, H. Zhou, S. H. Li, A. A. Geim, T. T. Wang, N. Maskaraet al., “High- fidelity parallel entangling gates on a neutral-atom quantum computer,” Nature, vol. 622, no. 7982, pp. 268–272, 2023

View on Pith 2023

[5] [5]

S. J. Evered, M. Xu, S. H. Li, A. A. Geim, J. P. Bonilla Ataides, M. Kalinowski, D. Bluvstein, N. Maskara, C. Kokail, M. Greineret al., “High-fidelity entangling gates and nonlocal circuits with neutral atoms,” Science, vol. 394, no. 6820, pp. 197–201, 2026

View on Pith 2026

[6] [6]

E. Knill and R. Laflamme, “Concatenated quantum codes,”arXiv preprint quant-ph/9608012, 1996

View on Pith Pith review arXiv 1996

[7] [7]

M. Grassl, P. Shor, G. Smith, J. Smolin, and B. Zeng, “Generalized concatenated quantum codes,”Physical Review A—Atomic, Molecular, and Optical Physics, vol. 79, no. 5, p. 050306, 2009. 0.01 0.015 0.02 0.03 0.05 0.07 0.10 10−4 10−3 10−2 10−1 100 Physical Error Ratep x Logical Error Rate (a) BP BP, BB [[144, 12]] BP, LA-BB [[144, 10]] HLLD-MAP, LA-BB [[144...

View on Pith 2009

[8] [8]

M. J. Gullans, S. Krastanov, D. A. Huse, L. Jiang, and S. T. Flammia, “Quantum coding with low-depth random circuits,”Physical Review X, vol. 11, no. 3, p. 031066, 2021

View on Pith 2021

[9] [9]

C. Gidney, M. Newman, P. Brooks, and C. Jones, “Yoked surface codes,” Nature Communications, vol. 16, no. 1, p. 4498, 2025

View on Pith 2025

[10] [10]

P. Shanahan and D. Ruiz, “Elevator codes: Concatenation for resource-efficient quantum memory under biased noise,”arXiv preprint arXiv:2601.10786, 2026

View on Pith Pith review arXiv 2026

[11] [11]

P. Panteleev and G. Kalachev, “Degenerate quantum LDPC codes with good finite length performance,”Quantum, vol. 5, p. 585, 2021

View on Pith 2021

[12] [12]

H. Yao, W. A. Laban, C. Häger, A. G. i Amat, and H. D. Pfister, “Belief propagation decoding of quantum LDPC codes with guided decimation,” in2024 IEEE International Symposium on Information Theory (ISIT). IEEE, 2024, pp. 2478–2483

View on Pith 2024

[13] [13]

J. Du Crest, M. Mhalla, and V . Savin, “Stabilizer inactivation for message- passing decoding of quantum LDPC codes,” in2022 IEEE Information Theory Workshop (ITW). IEEE, 2022, pp. 488–493

View on Pith 2022

[14] [14]

S. Koutsioumpas, H. Sayginel, M. Webster, and D. E. Browne, “Auto- morphism ensemble decoding of quantum LDPC codes,”arXiv preprint arXiv:2503.01738, 2025

View on Pith Pith review arXiv 2025

[15] [15]

S. Rabeti and H. Mahdavifar, “Multiple-bases belief propagation list decoding for quantum LDPC codes,” 2026, accepted to IEEE Information Theory Workshop (ITW 2026), available at arXiv:2605.14170

View on Pith Pith review arXiv 2026

[16] [16]

D. Poulin, “Optimal and efficient decoding of concatenated quantum block codes,”Physical Review A—Atomic, Molecular, and Optical Physics, vol. 74, no. 5, p. 052333, 2006

View on Pith 2006

[17] [17]

C. Zhang, Z. Wu, J. Wu, and S. Huang, “Bidirectional decoding for con- catenated quantum hamming codes,”arXiv preprint arXiv:2601.09131, 2026

View on Pith arXiv 2026

[18] [18]

T. Kakizaki, “Approximate level-by-level maximum-likelihood decoding based on the chase algorithm for high-rate concatenated stabilizer codes,” arXiv preprint arXiv:2601.18743, 2026

View on Pith arXiv 2026

[19] [19]

S. Kudekar, T. Richardson, and A. Iyengar, “The effect of saturation on belief propagation decoding of LDPC codes,” in2014 IEEE International Symposium on Information Theory. IEEE, 2014, pp. 2604–2608

View on Pith 2014

[20] [20]

F. J. MacWilliams and N. J. A. Sloane,The theory of error-correcting codes. Elsevier, 1977, vol. 16

View on Pith 1977

[21] [21]

J. Roffe, “LDPC: Python tools for low density parity check codes,”

View on Pith

[22] [22]

J. Roffe, D. R. White, S. Burton, and E. Campbell, “Decoding across the quantum low-density parity-check code landscape,”Physical Review Research, vol. 2, no. 4, Dec. 2020

View on Pith 2020

[23] [23]

M. Cain, Q. Xu, R. King, L. R. Picard, H. Levine, M. Endres, J. Preskill, H.-Y . Huang, and D. Bluvstein, “Shor’s algorithm is possible with as few as 10,000 reconfigurable atomic qubits,”arXiv preprint arXiv:2603.28627, 2026

View on Pith Pith review arXiv 2026

[24] [24]

using a repetition LPC of lengthr= 3: (a) BP and (b) BP-OSD, comparing direct decoding with the proposed HLLD-MAP and HLLD-FULL strategies. 0.02 0.025 0.03 0.04 0.05 10−4 10−3 10−2 10−1 Physical Error Ratep x Logical Error Rate BP, Mitten [[780, 156]] BP, LA-Mitten [[780, 147]] HLLD-MAP, LA-Mitten [[780, 147]] HLLD-FULL, LA-Mitten [[780, 147]] Fig. 4: Dec...

View on Pith

[25] [25]

using a repetition LPC of lengthr= 10, under direct BP and the proposed HLLD-MAP and HLLD-FULL

View on Pith

[26] [26]

A. A. Kovalev and L. P. Pryadko, “Quantum kronecker sum-product low- density parity-check codes with finite rate,”Phys. Rev. A, vol. 88, p. 012311, Jul 2013

View on Pith 2013

[27] [27]

S. Bravyi, A. W. Cross, J. M. Gambetta, D. Maslov, P. Rall, and T. J. Yoder, “High-threshold and low-overhead fault-tolerant quantum memory,”Nature, vol. 627, no. 8005, p. 778–782, Mar. 2024

View on Pith 2024

[28] [28]

A. Bhardwaj, M. Ma, N. Meister, R. King, D. Bluvstein, J. Preskill, M. Cain, Q. Xu, and H.-Y . Huang, “High-rate qLDPC processors,”arXiv preprint arXiv:2607.28795, 2026

View on Pith Pith review arXiv 2026

[29] [30]

R. Wang and L. P. Pryadko, “Distance bounds for generalized bicycle codes,” 2022. [Online]. Available: https://arxiv.org/abs/2203.17216

View on Pith Pith review arXiv 2022

[30] [31]

S. Rabeti and H. Mahdavifar, “Univariate bicycle quantum LDPC codes: Explicit logical structure and distance bounds,” 2026, accepted to IEEE Information Theory Workshop (ITW 2026), available at arXiv:2605.14173

View on Pith Pith review arXiv 2026

[31] [2022]

[Online]. Available: https://pypi.org/project/ldpc/

View on Pith