REVIEW 3 major objections 3 minor 25 references
CSS Quantum LRCs with Intersecting Recovery Sets: Constructions and Bounds
T0 review · 3 major / 3 minor · reviewed 2026-08-12 · deepseek-v4-flash
Pith's one-line read This paper proves that CSS quantum locally recoverable codes with intersecting recovery sets are exactly classical LRCs with common recovery sets, constructs binary families from subset-inclusion matrices, and derives a Singleton-like…
desk verdict Clean iff characterization and a good subset-inclusion construction for CSS qLRCs; the exact-case Singleton bound is a proof sketch resting on an unstated lemma and should not be relied on as written. read the letter →
The pith
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
The reading
What carries the argument
Two mechanisms carry the argument. First, the CSS local-recovery criterion is translated through the assumption $d(C^\perp)\ge 2$, which makes the shortened-code condition $\sigma_{\{i\}}(\pi_{R}(C))=\{0\}$ reduce to the same condition for both classical codes; this is why the phrase 'common recovery sets' appears. Second, the subset-inclusion matrix $H_{m,s,\alpha}$ over $\mathbb{F}_2$, with rows indexed by $(s-\alpha)$-subsets of $[m]$ and columns by $s$-subsets, entry 1 iff the row set is contained in the column set, plays the role of parity-check matrix for a binary code $C_{m,s,\alpha}$. The support of each row is a recovery set, giving locality $r=\binom{m-s+\alpha}{\alpha}-1$, availability $t=\binom{s}{\alpha}$, and intersection parameter $x=\binom{m-s+\alpha-1}{\alpha-1}-1$. Dual-containment is characterized by $H_{m,s,\alpha}H_{m,s,\alpha}^T=0$, i.e. by binomial congruence conditions, which enables the CSS construction.
What would settle it
Enumerate all binary exact $(r,t,x)$-cLRCs for small $n$ and check directly whether the ordering required by Lemma 14 exists; exhibit one exact code where no such ordering exists, and the Singleton-like bound in Corollary 4 has no proof. Similarly, an explicit CSS code pair with $d(C^\perp)\ge 2$ that is an $(r,t,x)$-qLRC but whose classical codes are not $(r,t,x)$-cLRCs with common recovery sets would refute Theorem 1.
Extended reading notes
Core claim
On the paper's own terms, the central discovery is a bridge between the quantum and classical world: the local-recoverability structure of a CSS code with intersecting recovery sets is captured exactly by classical codes, except for a mild nondegeneracy condition. Theorem 1 states that for $C_1,C_2$ with $d(C_1^\perp),d(C_2^\perp)\ge 2$ and $C_1^\perp\subseteq C_2$, the code $\mathrm{CSS}(C_1,C_2)$ is an $(r,t,x)$-qLRC if and only if $C_1$ and $C_2$ are $(r,t,x)$-cLRCs with common recovery sets. The proof reduces the quantum recovery conditions to shortened-code conditions that become empty exactly because the dual distance is at least 2. The paper then exhibits the subset-inclusion codes $C_{m,s,\alpha}$ as $(r,t,x)$-cLRCs with explicit locality, availability, and intersection parameter, characterizes when they are dual-containing by parity conditions on binomial coefficients, and obtains binary CSS $(r,t,x)$-qLRCs. In the exact case, it proves a Singleton-like bound $\kappa\le n-2(d-1)-2\bar{N}$, and constructs an exact pure family with parameters $[[\binom{m}{3},\binom{m}{3}-2m,4]]$ whose rate approaches 1.
Load-bearing premise
The load-bearing premise is the imported Lemma 14 of [9], which guarantees that in every exact $(r,t,x)$-cLRC the coordinates can be ordered so that each coordinate's chosen recovery set avoids all previously erased coordinates; Proposition 2 is only sketched and would collapse if that lemma does not transfer to classical exact codes.
Editorial extensions
If this is right
- Any classical $(r,t,x)$-cLRC with common recovery sets and dual minimum distance at least 2 immediately yields a CSS $(r,t,x)$-qLRC with the same recovery sets, so classical construction techniques become quantum construction techniques under a checkable condition.
- The subset-inclusion matrix family provides infinite families of binary CSS $(r,t,x)$-qLRCs with explicit parameter formulas and, for the parameter choices in Table I, high rates between 0.45 and 0.86 with guaranteed distances up to 16.
- The exact $(3,2)$-subfamily is pure with distance 4, dimension $\binom{m}{3}-2m$, and rate $1-\frac{12}{(m-1)(m-2)}$, which tends to 1; it offers a rate advantage over the only previously known explicit exact construction while giving up on small intersection parameter.
- The CSS-specific Singleton-like bound $\kappa\le n-2(d-1)-2\bar{N}$ matches the general exact bound of [9] for the exact subfamily when $m\ge 14$.
- Because the exact subfamily is pure, its quantum minimum distance equals the classical distance rather than merely being lower-bounded by it, so the distance bound from Corollary 3 applies directly to the quantum code.
Reading between the lines
- Beyond the paper's claims, the parity self-orthogonality criterion for incidence matrices likely generalizes to other regular set systems, such as subspace-inclusion matrices over $\mathbb{F}_q$, giving non-binary CSS $(r,t,x)$-qLRCs; the paper explicitly leaves the non-binary extension as future work.
- The common-recovery-set obstruction suggests that optimizing the intersection parameter $x$ rather than minimizing it may be the right quantum regime: $x=0$ is impossible, and the examples show high rate comes with large $x$. One could test whether constructions with the smallest possible $x=1$ necessarily have vanishing rate, refining the tradeoff seen in [9].
- Because the distance is fixed at 4 in the exact family while the upper bound grows with $m$, the construction is likely far from optimal for large distance; pushing the inclusion-matrix parameters to get distance scaling would settle whether subset-inclusion codes can approach the distance bound.
- The exact-case bound relies on a lemma imported from [9]. A small exhaustive search over exact $(r,t,x)$-cLRCs would show whether the ordering lemma actually holds in all binary cases, which would confirm or falsify whether the bound transfers cleanly.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The paper studies CSS quantum locally recoverable codes with multiple intersecting recovery sets, the (r,t,x)-qLRC setting of Bu, Gu, and Li. Its stated contributions are threefold: (i) an iff characterization, under a dual-minimum-distance condition, of CSS (r,t,x)-qLRCs in terms of classical (r,t,x)-cLRCs with common recovery sets; (ii) an explicit binary subset-inclusion construction of dual-containing (r,t,x)-cLRCs, with a complete binomial condition for dual-containment and an exact (3,2)-family with parameters [[C(m,3), C(m,3)-2m, 4]]; and (iii) bounds for CSS (r,t,x)-qLRCs, including a Singleton-like dimension bound in the exact case and a comparison of the exact family with bounds and with the earlier Bu-Gu-Li construction.
Significance. If the paper's claims are correct, the main positive contribution is substantial: Theorem 1 gives a clean and useful bridge between classical and quantum locality for intersecting recovery sets, and the subset-inclusion construction of Theorems 2-3 is explicit, with a checkable dual-containment congruence and an exact infinite family whose rate approaches 1 at fixed distance 4. I found no fitted parameters and no circular reasoning: the construction is explicit and the bounds are substitutions of cited classical bounds. However, the exact-case Singleton-like bound, which is one of the advertised contributions and supports the Section VI comparison, is currently not actually proved in the manuscript, and Table II contains numerical values that do not follow from the formulas as written. These are load-bearing issues that require a major revision.
major comments (3)
- [V-B, Proposition 2] The exact-case Singleton-like bound is not proved in this manuscript. The proof of Proposition 2 is a four-step sketch whose key step is an appeal to '[9, Lemma 14]', a lemma that is neither stated nor proved here, and no argument is given that this lemma transfers from the quantum setting of [9] to classical exact (r,t,x)-cLRCs. The quantity \bar N(n,r,d,⌈n p_e⌉) is also asserted without derivation. Because Corollary 4 and the Section VI comparison depend entirely on Proposition 2, the authors should either state and prove the needed lemma (or give a self-contained derivation of the ordering and counting arguments) and derive \bar N, or remove the affected claims.
- [VI, Table II] The m=10 row of Table II is inconsistent with the formulas in Corollary 4 as written. With n=120, r=35, x=7, t=3, d=4, the stated p_e is 9/72 - 9/128 + 1/170 = 0.06057, so ⌈n p_e⌉=8; the condition in \bar N becomes 36N - (N(N-1)/56)·168 ≤ 117, whose largest solution is N=4; Corollary 4 then gives κ ≤ 120 - 2(3) - 2·4 = 106, not the listed 104. I also find discrepancies for m=18,22,26 (the formula gives 792,1512,2566 respectively, while the table lists 790,1508,2564). The table and the formulas must be reconciled before the comparison in Section VI can be used as evidence.
- [V-B, Proposition 2 and Definition 4] Even granting the existence of the unstated Lemma 14, the proof sketch of Proposition 2 requires an ordering i_1,...,i_T such that for every ℓ≥d some recovery set of i_ℓ avoids all earlier coordinates. Definition 4 only controls intersections among the t recovery sets of a single coordinate; it imposes no visible bound on how often one coordinate belongs to another coordinate's recovery sets. The ordering property is therefore nontrivial and must be proved (or derived explicitly from the cited lemma) before the puncturing argument k ≤ n - |U| can be accepted.
minor comments (3)
- [III, Theorem 1] The iff characterization is correctly stated as conditional on d(C⊥_ℓ)≥2, but the paper would benefit from a sentence in the conclusion noting that this assumption excludes CSS codes whose duals have weight-one codewords; in that excluded case Lemma 1 would give non-vanishing local syndromes rather than the classical condition.
- [VI, Table II] The notation κ~_e^* is used in the table and caption but the tilde is never defined; the caption should explicitly say that κ~_e^* denotes the CSS exact bound from Corollary 4.
- [IV-B, Corollary 2] The sentence 'Since each column of H_{m,s,α} has weight C(s,α)>0, C_{m,s,α} has no weight-one codeword' is correct but would be clearer if it noted explicitly that a weight-one codeword e_i would require the i-th column of H to be zero.
Circularity Check
No significant circularity; the derivation chain is explicit and rests on external, non-fitted inputs.
full rationale
The paper's derivation chain is self-contained. Theorem 1 is an iff statement whose reduction step is Lemma 1, imported from the external Galindo et al. criterion [4, Prop. 26]; the assumption d(C^⊥)≥2 turns the CSS recovery condition σ_i(π_R(C_l)) = σ_i(C^⊥_{3-l}) into the zero condition that defines a classical recovery set, so the equivalence is a genuine reduction, not a definitional tautology. The construction is explicit: H_{m,s,α} is a concrete subset-inclusion matrix, and the parameters r, t, and x are computed by direct combinatorial counts (number of rows through a column, and intersection size of two row supports), with dual-containment characterized by the binomial congruence in Theorem 3; no fitted parameter is later renamed as a prediction. The CSS parameters follow from the standard CSS formula and Theorem 1. The bounds in Corollary 3 are substitutions of the external classical bounds of Kruglik et al. [11], [12], and Corollary 4 is an algebraic rearrangement of Proposition 2. The only load-bearing weakness is Proposition 2's proof sketch, which invokes the unstated [9, Lemma 14] to produce the ordering i_1,...,i_T and the quantity ar N; that is a proof gap and a correctness risk if the lemma or the count does not transfer, but it is not circularity because [9] is an independent external work rather than a self-citation, and the bound is not assumed as an input. The self-citation [2] (Ramkumar is a coauthor) appears only in a background listing and supports no central claim. No fitted input is called a prediction, and no uniqueness theorem from the authors' own prior work is imported to force a choice.
Assumptions & free parameters
assumptions (6)
- domain assumption CSS local erasure criterion of Galindo et al. [4, Prop. 26]
- domain assumption Classical (r,t,x)-cLRC dimension bound k ≤ n(1-p(r,t,x)) from [11]
- domain assumption Alphabet-dependent distance bound for (r,t,x)-cLRCs from [12]
- domain assumption Dimension formula and distance bounds for subset-inclusion codes from Wilson [23] and Marin-Mogilnykh [24]
- domain assumption Lemma 14 of Bu-Gu-Li [9] on recovery-set structure in exact codes
- domain assumption Dual minimum distance at least two for underlying classical codes
Cite this review
Pith. "Pith review of CSS Quantum LRCs with Intersecting Recovery Sets: Constructions and Bounds." pith.science (2026). https://pith.science/paper/MQE5ZMKH
@misc{pith2026260810912,
author = {Pith},
title = {Pith review of: CSS Quantum LRCs with Intersecting Recovery Sets: Constructions and Bounds},
year = {2026},
howpublished = {\url{https://pith.science/paper/MQE5ZMKH}},
note = {Machine review of arXiv:2608.10912}
}
abstract
In this work, we study $(r,t,x)$ quantum locally recoverable codes (qLRCs) with locality $r$, $t$ recovery sets per qudit, and intersection parameter $x$. We first show that, assuming the underlying classical codes have dual minimum distance at least two, a CSS code is an $(r,t,x)$-qLRC if and only if the underlying classical codes are $(r,t,x)$ classical LRCs (cLRCs) with common recovery sets. We then use subset-inclusion matrices to construct families of binary dual-containing $(r,t,x)$-cLRCs, which yield binary $(r,t,x)$-qLRCs via the CSS construction. For CSS $(r,t,x)$-qLRCs, we derive upper bounds on the dimension and rate, minimum-distance bounds in the pure case, and a Singleton-like dimension bound in the exact case. Finally, we show that these families attain high rates and nontrivial minimum distances.
Reference graph
Works this paper leans on
-
[9]
Quantum locally recoverable code with intersecting recovery sets,
K. Bu, W. Gu, and X. Li, “Quantum locally recoverable code with intersecting recovery sets,”arXiv preprint arXiv:2501.10354, 2025
arXiv 2025
-
[1]
Quantum locally recoverable codes,
L. Golowich and V . Guruswami, “Quantum locally recoverable codes,” arXiv preprint arXiv:2311.08653, 2023
arXiv 2023
-
[2]
Quantum locally recoverable codes via good polynomials,
S. Sharma, V . Ramkumar, and I. Tamo, “Quantum locally recoverable codes via good polynomials,”IEEE Journal on Selected Areas in Information Theory, 2025
2025
-
[3]
Bounds and constructions of quantum locally recoverable codes from quantum CSS codes,
G. Luo, B. Chen, M. F. Ezerman, and S. Ling, “Bounds and constructions of quantum locally recoverable codes from quantum CSS codes,”IEEE Transactions on Information Theory, 2025
2025
-
[4]
Quantum (r,δ)-locally recoverable codes,
C. Galindo, F. Hernando, H. Mart ´ın-Cruz, and R. Matsumoto, “Quantum (r,δ)-locally recoverable codes,”Finite Fields and Their Applications, vol. 111, p. 102785, 2026
2026
-
[5]
Optimal quantum(r, δ)-locally repairable codes via classical ones,
K. Zhou and M. Cao, “Optimal quantum(r, δ)-locally repairable codes via classical ones,”arXiv preprint arXiv:2507.18175, 2025
arXiv 2025
-
[6]
Two families of optimal quantum locally recoverable codes,
D. Xie, S. Zhu, and Z. Sun, “Two families of optimal quantum locally recoverable codes,”International Journal of Theoretical Physics, vol. 64, no. 4, pp. 1–17, 2025
2025
-
[7]
Improved bounds and optimal constructions of pure quantum locally recoverable codes,
Y . Li, S. Li, G. Luo, and S. Ling, “Improved bounds and optimal constructions of pure quantum locally recoverable codes,”arXiv preprint arXiv:2512.07256, 2025
arXiv 2025
Show all 25 references
-
[8]
Optimal quantum(r, δ)-locally repairable codes from matrix-product codes,
M. Cao and K. Zhou, “Optimal quantum(r, δ)-locally repairable codes from matrix-product codes,”arXiv preprint arXiv:2508.03597, 2025
2025 arXiv
-
[10]
Locality and availability in distributed storage,
A. S. Rawat, D. S. Papailiopoulos, A. G. Dimakis, and S. Vishwanath, “Locality and availability in distributed storage,”IEEE Transactions on Information Theory, vol. 62, no. 8, pp. 4481–4493, 2016
2016
-
[11]
On one general- ization of lrc codes with availability,
S. Kruglik, M. Dudina, V . Potapova, and A. Frolov, “On one general- ization of lrc codes with availability,” in2017 IEEE Information Theory Workshop (ITW), pp. 26–30, IEEE, 2017
2017
-
[12]
On distance properties of(r, t, x)-lrc codes,
S. Kruglik, K. Nazirkhanova, and A. Frolov, “On distance properties of(r, t, x)-lrc codes,” in2018 IEEE International Symposium on Information Theory (ISIT), pp. 1336–1339, IEEE, 2018
2018
-
[13]
New bounds and general- izations of locally recoverable codes with availability,
S. Kruglik, K. Nazirkhanova, and A. Frolov, “New bounds and general- izations of locally recoverable codes with availability,”IEEE Transac- tions on Information Theory, vol. 65, no. 7, pp. 4156–4166, 2019
2019
-
[14]
Bounds on the parameters of locally recoverable codes,
I. Tamo, A. Barg, and A. Frolov, “Bounds on the parameters of locally recoverable codes,”IEEE Transactions on information theory, vol. 62, no. 6, pp. 3070–3083, 2016
2016
-
[15]
Achieving arbitrary locality and availability in binary codes,
A. Wang, Z. Zhang, and M. Liu, “Achieving arbitrary locality and availability in binary codes,” in2015 IEEE International Symposium on Information Theory (ISIT), pp. 1866–1870, IEEE, 2015
2015
-
[16]
Constructions of binary locally repairable codes with multiple recovering sets,
J. Teng and L. Jin, “Constructions of binary locally repairable codes with multiple recovering sets,”IEEE Access, vol. 9, pp. 92239–92245, 2021
2021
-
[17]
Two classes of (r, t)-locally repairable codes,
A. Wang, Z. Zhang, and D. Lin, “Two classes of (r, t)-locally repairable codes,” in2016 IEEE International Symposium on Information Theory (ISIT), pp. 445–449, IEEE, 2016
2016
-
[18]
RS-like locally recoverable codes with intersecting recovering sets,
C. Rajput and M. Bhaintwal, “RS-like locally recoverable codes with intersecting recovering sets,”Finite Fields and Their Applications, vol. 68, p. 101729, 2020
2020
-
[19]
A subclass of lrc codes with intersecting recovering sets,
C. Rajput and M. Bhaintwal, “A subclass of lrc codes with intersecting recovering sets,” in2022 IEEE Information Theory Workshop (ITW), pp. 512–516, IEEE, 2022
2022
-
[20]
On the locality of codeword symbols,
P. Gopalan, C. Huang, H. Simitci, and S. Yekhanin, “On the locality of codeword symbols,”IEEE Transactions on Information theory, vol. 58, no. 11, pp. 6925–6934, 2012
2012
-
[21]
Quantum error correction via codes over gf (4),
A. R. Calderbank, E. M. Rains, P. M. Shor, and N. J. Sloane, “Quantum error correction via codes over gf (4),”IEEE Transactions on Informa- tion Theory, vol. 44, no. 4, pp. 1369–1387, 1998
1998
-
[22]
Gottesman,Stabilizer Codes and Quantum Error Correction
D. Gottesman,Stabilizer Codes and Quantum Error Correction. PhD thesis, California Institute of Technology, 1997
1997
-
[23]
A diagonal form for the incidence matrices of t-subsets vs. k-subsets,
R. M. Wilson, “A diagonal form for the incidence matrices of t-subsets vs. k-subsets,”European Journal of Combinatorics, vol. 11, no. 6, pp. 609–615, 1990
1990
-
[24]
Binary codes from subset inclusion matrices,
A. D. Marin and I. Y . Mogilnykh, “Binary codes from subset inclusion matrices,”Journal of Combinatorial Designs, vol. 34, no. 2, pp. 87–103, 2026
2026
-
[25]
On optimal quantum LRCs from the Hermitian construction andt-designs,
Y . Li, S. Li, H. Lao, G. Luo, and S. Ling, “On optimal quantum LRCs from the Hermitian construction andt-designs,”arXiv preprint arXiv:2508.13553, 2025
2025 arXiv
Reviewed August 12, 2026 · model on record in the stance chip above.
Discussion (0). Continue with ORCID to comment.