REVIEW
Random Quantum LDPC Codes Approaching the Gilbert-Varshamov Bound
Information Theory · Submitted 2 Oct 2026 · 27 references
Pith's verdict: The η-degeneracy machinery is the real contribution and is genuinely new, but the full rate range of the main theorem rests on a base-code existence claim the cited papers do not support.
T0 · AI review
Why Pith says this
Highlight any text to comment on it.
Referee report
The paper
The paper claims
This paper claims that quantum LDPC codes can match the best rate–distance tradeoff of arbitrary stabilizer codes: for any prime p, rate R, and ε > 0, there is an ensemble of p-ary QLDPC codes of rate R and constant-weight checks whose…
The reading
What carries the argument
The central object is the η-degenerate QLDPC outer code produced by the Alon–Edmonds–Luby (AEL) distance amplification: take a high-rate asymptotically good QLDPC base code, encode each of its coordinates with one short pure Reed–Solomon inner code, and fold the result along the edges of a bipartite expander. The expander mixing lemma (and the purity of the inner code) forces the low-weight vectors of the folded stabilizer space to be few — at most |L|^η of them below the distance — which is exactly the quantitative handle needed to make the random-concatenation union bound non-vacuous. This relaxed degeneracy bound replaces the purity of Reed–Solomon outer codes that Ouyang's earlier GV-bound argument relied on, and it is what turns the random concatenation into an LDPC ensemble that still looks random in the pointwise undetectability sense.
Extended reading notes
Core claim
The central claim is Theorem 5.8: for any prime p, any rational rate ρ ∈ (0,1), and any ε > 0, there is an ensemble of p-ary QLDPC codes of rate ρ and locality O(1) such that, with probability 1 − exp(−Ω(n)), a random code has relative distance exceeding $H^{{-1}}$_{$p^{2}$}((1−ρ)/2) − ε, i.e., the quantum GV bound minus ε. The discovery that makes this possible is quantitative control of degeneracy. Because a QLDPC code's stabilizer space is generated by constant-weight checks, it always contains low-weight vectors; the paper defines a code to be η-degenerate when fewer than |L|^η of its stabilizers have weight below its distance, and proves (Theorem 1.5) that quantum AEL codes achieve distance arbitrarily close to the quantum Singleton bound (1−R−ε)/2 while being η-degenerate for any fixed small η. This η-degeneracy keeps the union-bound count non-vacuous and converts a Thommesen-style random concatenation — Ouyang's construction with Reed–Solomon outer codes replaced by these LDPC outer codes — into an ensemble that behaves like a random stabilizer code pointwise: for any fixed error e, the probability that e is undetectable is at most $p^{{−N(1−R−ε)}}$. That single pointwise bound yields both the GV distance and the channel-capacity statements (Theorem 1.3).
Load-bearing premise
The construction assumes there is a family of p-ary QLDPC base codes whose rate can be pushed to 1−γ for arbitrarily small γ while their relative distance stays an absolute positive constant δ_base; the AEL distance step δ_AEL ≥ α − λ/δ_base collapses if δ_base shrinks with γ, because the folded outer code would then lose the distance the GV argument needs.
Editorial extensions
If this is right
- QLDPC codes achieve the same rate–distance tradeoff as arbitrary stabilizer codes: low density costs nothing asymptotically at the GV bound.
- The ensemble reaches rate 1−2κ−ε on the quantum erasure channel and the hashing bound on memoryless Pauli channels (including depolarizing), matching random stabilizer codes up to an arbitrary ε-loss.
- The pointwise undetectability bound yields a QLDPC purity-testing family that admits efficient decoding, a feature prior purity-testing constructions (curve- or code-based) were not known to have.
- In the general-inner-code variant, the concatenated codes carry a certifiable worst-case distance (product of outer and inner distances) and a decoder correcting a constant fraction of errors, properties generic random stabilizer codes lack.
- Because the outer locality and inner blocklength are constant, the whole family is QLDPC with O_ε(1) check weight, making the GV distance approachable by a structured, not merely existential, family.
Reading between the lines
- The template — a rigid global LDPC skeleton with local random inner codes reproducing macroscopic randomness — suggests a general principle that may transfer to other channels or parameter regimes where random stabilizer codes are known to be tight, not just the three settings treated here.
- The η-degeneracy parameter is held constant in the proof; pushing it toward the check-weight lower bound of Observation 5.1 (weight Ω(1/η)) would interpolate toward a more uniformly random ensemble, possibly capturing channel properties beyond the pointwise undetectability bound.
- The result is probabilistic and does not give explicit (non-random) QLDPC codes at the GV bound; since no explicit classical codes match GV either, explicit constructions remain open.
- The erasure and Pauli-channel results carry the same γ-loss as the pointwise bound; whether the loss can be removed entirely (exact capacity rather than capacity−ε) is left open.
Original abstract
We show that for any $R,\epsilon >0$ and any prime $p\geq 2$, there exists an infinite family of $p$-ary quantum low-density parity-check (QLDPC) codes, rate $R$, checks of weight $O_\epsilon(1)$, and normalized distance at least $\delta_{\mathrm{GV}}(p,R)-\epsilon$. Here, $\delta_{\mathrm{GV}}(p,R)$ denotes the quantum Gilbert-Varshamov (GV) bound for $p$-ary stabilizer codes. In fact, we construct an ensemble of such QLDPC codes, such that a random code from this ensemble is close to the GV bound with high probability. Moreover, this ensemble matches the performance of random stabilizer codes on several quantum channels. Specifically, it approaches the quantum capacity of the erasure channel and the hashing bound for memoryless Pauli channels, including the depolarizing channel. A significant challenge in working with QLDPC codes is that they are necessarily \emph{degenerate}, i.e., contain many low-weight stabilizers. A key contribution of our work is a construction of QLDPC codes with quantitative control on their degeneracy. These codes are obtained by combining known constructions of asymptotically good QLDPC codes with the expander-based distance amplification procedure of Alon, Edmonds, and Luby [FOCS'95]. Our random ensemble is constructed by starting with these low-degeneracy QLDPC codes near the quantum Singleton bound and concatenating each coordinate with a random inner code. This can be viewed as a quantum analogue of Thommesen's construction, and as an LDPC version of a result of Ouyang, with an appropriately designed outer code.
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.
The abstract, sentence by sentence
We show that for any $R,\epsilon >0$ and any prime $p\geq 2$, there exists an infinite family of $p$-ary quantum low-density parity-check (QLDPC) codes, rate $R$, checks of weight $O_\epsilon(1)$, and normalized distance at least $\delta_{\mathrm{GV}}(p,R)-\epsilon$. Here, $\delta_{\mathrm{GV}}(p,R)$ denotes the quantum Gilbert-Varshamov (GV) bound for $p$-ary stabilizer codes. In fact, we construct an ensemble of such QLDPC codes, such that a random code from this ensemble is close to the GV bound with high probability. Moreover, this ensemble matches the performance of random stabilizer codes on several quantum channels. Specifically, it approaches the quantum capacity of the erasure channel and the hashing bound for memoryless Pauli channels, including the depolarizing channel. A significant challenge in working with QLDPC codes is that they are necessarily \emph{degenerate}, i.e., contain many low-weight stabilizers. A key contribution of our work is a construction of QLDPC codes with quantitative control on their degeneracy. These codes are obtained by combining known constructions of asymptotically good QLDPC codes with the expander-based distance amplification procedure of Alon, Edmonds, and Luby [FOCS'95]. Our random ensemble is constructed by starting with these low-degeneracy QLDPC codes near the quantum Singleton bound and concatenating each coordinate with a random inner code. This can be viewed as a quantum analogue of Thommesen's construction, and as an LDPC version of a result of Ouyang, with an appropriately designed outer code.
background contribution method result interpretation main claim
| # | sentence | role | the authors' own work | states the main conclusion | strength |
|---|---|---|---|---|---|
| s0 | We show that for any $R,\epsilon >0$ and any prime $p\geq 2$, there exists an infinite family of $p$-ary quantum low-density parity-check (QLDPC) codes, rate $R$, checks of weight $O_\epsilon(1)$, and normalized distance at least $\delta_{\mathrm{GV}}(p,R)-\epsilon$. | result 0.89 | 0.96 | 0.89 | asserted 1.00 |
| s1 | Here, $\delta_{\mathrm{GV}}(p,R)$ denotes the quantum Gilbert-Varshamov (GV) bound for $p$-ary stabilizer codes. | background 0.96 | 0.12 | 0.04 | asserted 0.88 |
| s2 | In fact, we construct an ensemble of such QLDPC codes, such that a random code from this ensemble is close to the GV bound with high probability. | contribution 0.59 | 0.95 | 0.24 | asserted 0.99 |
| s3 | Moreover, this ensemble matches the performance of random stabilizer codes on several quantum channels. | result 0.97 | 0.90 | 0.10 | asserted 1.00 |
| s4 | Specifically, it approaches the quantum capacity of the erasure channel and the hashing bound for memoryless Pauli channels, including the depolarizing channel. | result 0.97 | 0.83 | 0.09 | asserted 0.99 |
| s5 | A significant challenge in working with QLDPC codes is that they are necessarily \emph{degenerate}, i.e., contain many low-weight stabilizers. | background 0.96 | 0.07 | 0.02 | asserted 1.00 |
| s6 | A key contribution of our work is a construction of QLDPC codes with quantitative control on their degeneracy. | contribution 1.00 | 0.95 | 0.11 | asserted 1.00 |
| s7 | These codes are obtained by combining known constructions of asymptotically good QLDPC codes with the expander-based distance amplification procedure of Alon, Edmonds, and Luby [FOCS'95]. | method 0.98 | 0.51 | 0.04 | asserted 1.00 |
| s8 | Our random ensemble is constructed by starting with these low-degeneracy QLDPC codes near the quantum Singleton bound and concatenating each coordinate with a random inner code. | method 0.99 | 0.90 | 0.04 | asserted 1.00 |
| s9 | This can be viewed as a quantum analogue of Thommesen's construction, and as an LDPC version of a result of Ouyang, with an appropriately designed outer code. | interpretation 0.78 | 0.58 | 0.04 | asserted 0.84 |
What was examined, and what was not
- arXiv listing abstract v1
- full text
- figures and tables
- references and citing sentences
- relations to other papers
- P abstract
- S sections
- T body sentences
- C citations
- R cross-paper relations
Figures
Figures reproduced from arXiv:2610.02648 by the authors.
References
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.