Pith. sign in

REVIEW 3 major objections 2 minor 41 references

Many magic-friendly triples yield constant-depth CCZ gates

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

T0 review · deepseek-v4-flash

2026-08-04 06:13 UTC pith:FBXGEAXD

load-bearing objection Theorem 1's hypotheses are mutually inconsistent; the counting argument makes the main result vacuous. the 3 major comments →

arxiv 2601.22489 v2 pith:FBXGEAXD submitted 2026-01-30 quant-ph cs.ITmath.IT

Native Non-Clifford Gates in Quantum LDPC Codes: Conditions, Synthesis, and Scaling Limits

Mohammad Rowshan This is my paper
classification quant-ph cs.ITmath.IT MSC 81P7081P6805C65
keywords quantum LDPC codesnon-Clifford gatesCCZ gatesmagic statesconstant-depth circuitsCSS codestriple overlaphypergraph packing
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

This paper tries to establish a structural sufficient condition for building native non-Clifford gates on quantum LDPC codes. It defines 'magic-friendly triples' of logical X operators — three independent operators that are pairwise orthogonal and have an odd triple overlap — and shows that if a CSS code family on n qubits has at least Ω(n^{1+γ}) such triples with bounded per-qubit support participation, then a constant-depth circuit of physical CCZ gates implements Ω(n^γ) logical CCZ gates in parallel while preserving distance up to a constant factor. If true, this would turn the search for magic-state fountains on qLDPC codes into a combinatorial counting problem. The paper also argues that strict 'saturated' implementations face distance-depth tradeoffs that block constant-depth scaling, so scalable designs must relax strict subspace requirements or manage gate congestion. The result is a reduction, not a construction: it does not prove that existing good qLDPC families actually contain the needed triples.

Core claim

The central claim is Theorem 1: for a CSS code family on n qubits with distance Θ(n), if there exists a set of Ω(n^{1+γ}) magic-friendly triples — independent logical X operators with pairwise zero inner products and odd triple overlap — and each physical qubit participates in at most M of their supports, then there is a constant-depth circuit of physical CCZ gates that implements Ω(n^γ) logical CCZ gates in parallel, up to logical Clifford corrections, and the code distance remains Θ(n). The proof combines a packing lemma that selects a subcollection with disjoint supports, an edge-coloring lemma for bounded-degree 3-uniform hypergraphs that schedules the gates into constant depth, and a li

What carries the argument

The paper's machinery has four pieces: the magic-friendly triple (three independent logical X operators with pairwise orthogonality and odd triple overlap), which is the algebraic object carrying the logical CCZ phase; a 3-uniform hypergraph model of physical CCZ circuits, where vertices are physical qubits and hyperedges are CCZ gates, with maximum degree bounding circuit depth via edge coloring; a packing lemma that extracts a large subcollection of triples with pairwise disjoint supports from a bounded-participation pool; and a light-cone lemma showing constant-depth, bounded-locality circuits degrade code distance by at most a constant factor.

Load-bearing premise

The proof assumes, without proving, that every magic-friendly triple admits a bounded-locality diagonal circuit of physical Z, CZ, and CCZ gates that acts as a logical CCZ; this is non-obvious when the three logical supports overlap, because a physical CCZ gate cannot act on the same qubit multiple times.

What would settle it

Exhibit a magic-friendly triple (pairwise orthogonal, odd triple overlap) whose three supports overlap in a way that no bounded-locality diagonal circuit implements the logical CCZ; this would falsify the per-triple step. More directly, find a CSS code family with Ω(n^{1+γ}) magic-friendly triples of bounded participation for which no constant-depth physical CCZ circuit implements a superconstant number of logical CCZ gates while preserving distance.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • Any CSS code family satisfying the triple-count condition automatically has a native constant-depth CCZ magic-state fountain, removing the need for separate magic-state distillation in those codes.
  • The existence question for non-Clifford gates on asymptotically good qLDPC families is reduced to a concrete combinatorial problem: prove Ω(n^{1+γ}) magic-friendly triples with bounded participation exist in their logical X spaces.
  • Distance preservation up to a constant factor means the constant-depth CCZ construction is compatible with maintaining linear-distance codes.
  • Strict saturated (fully transversal) implementations cannot reach constant depth as code distance grows, so scalable designs must relax strict subspace conditions or manage gate congestion.
  • The same packing/coloring/light-cone recipe applies to other multi-controlled-Z magic-state factories, not just CCZ.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The load-bearing unproved step is the per-triple existence of a bounded-locality diagonal unitary implementing a logical CCZ from a magic-friendly triple; if that fails for overlapping supports, the constant-depth circuit construction collapses.
  • If the required triples are found in quantum Tanner codes or product-based qLDPC families, this would yield the first explicit qubit qLDPC family with constant rate, linear distance, and a native constant-depth magic-state fountain — a goal the paper states is open.
  • The combinatorial condition suggests a new target for high-dimensional expander theory: logical operators of good LDPC codes should be well-spread, and counting magic-friendly triples could be a testable proxy for that spreading.
  • A direct numerical test is possible: search for magic-friendly triples in small instances of known qLDPC codes and measure participation; if none scale as Ω(n^{1+γ}), the theorem's hypothesis is unlikely to hold for those families.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit.

Referee Report

3 major / 2 minor

Summary. The paper introduces 'magic-friendly triples' of X-type logical operators in CSS codes, characterized by pairwise orthogonality and odd triple overlap, and models physical CCZ circuits as bounded-degree 3-uniform hypergraphs. It claims a main structural theorem: if a CSS family has Ω(n^{1+γ}) such triples with linear-size supports and bounded per-qubit participation, then a constant-depth physical CCZ circuit implements Ω(n^γ) logical CCZ gates in parallel while preserving distance up to a constant factor. The proof combines a greedy packing lemma, an edge-coloring lemma, and a light-cone distance bound. The paper also gives small examples and frames the existence of such triples in asymptotically good qLDPC families as an open combinatorial problem.

Significance. If valid, the theorem would be a valuable reduction: it would turn the search for native CCZ magic-state fountains in qLDPC codes into a combinatorial counting problem on logical X operators, and the packing/coloring blueprint could apply to other multi-controlled-Z gates. The paper is clearly written and the authors honestly state that the hard combinatorial abundance question remains open. However, the central theorem is vacuous as stated because its hypotheses are mutually inconsistent, and the proof relies on an unproved per-triple gate-synthesis assertion. The claimed reduction is therefore not established; the framework may be salvageable only with substantial reformulation.

major comments (3)
  1. [Theorem 1 (Main structural theorem), conditions 2 and 3] Conditions 2 and 3 are inconsistent with condition 1. Since |S_t| ≥ a n for every t ∈ S_n and each coordinate belongs to at most M supports, double counting gives |S_n| a n ≤ Σ_t |S_t| = Σ_i |{t: i∈S_t}| ≤ M n, hence |S_n| ≤ M/a = O(1). But condition 1 demands |S_n| ≥ c_1 n^{1+γ} with γ>0. Thus no code family satisfies all three hypotheses for large n. The proof's Step 1 uses Lemma 2 to derive |T_n| ≥ (c_1/Mb)n^γ from |S_n| ≥ c_1 n^{1+γ}, but that premise cannot hold under the stated bounded-participation and linear-support assumptions. The advertised Ω(n^γ) fountain is not derived.
  2. [Appendix E, Step 2; also Remark 1 and Example 1] The proof assumes that every magic-friendly triple (x,y,z) automatically admits a bounded-locality diagonal unitary U_t implementing a logical CCZ. The text says one may take 'a transversal CCZ on physical representatives' and that 'the exact form is not important here, only the locality pattern,' but no construction or proof is given. This is load-bearing: the union of all U_t is the entire circuit. Moreover, Example 1 itself shows that the supports of x,y,z may overlap (qubit 4), in which case a 'transversal CCZ on representatives' is not a well-defined product of physical CCZ gates, since a physical CCZ acts on three distinct qubits. The paper also does not address the effect on spectator logical qubits when a logical operator participates in multiple triples. Without a rigorous per-triple lemma, the multi-triple circuit construction collapses.
  3. [Abstract and Contributions] The abstract and opening paragraph state that the paper establishes 'fundamental limits on code scaling' and 'tight distance-depth bounds' for strict saturated implementations. I could not find any theorem, lemma, or proof in the body corresponding to these negative claims. The only distance statement is the positive Lemma 4 (distance preserved up to a constant factor). If such no-go bounds are claimed, they need to be stated and proved; otherwise the abstract should be revised to match the content.
minor comments (2)
  1. [Appendix D, Lemma 4] The proof refers to the 'weight' of E = U†E' U, but E is not a Pauli operator when U is the non-Clifford CCZ circuit. The light-cone argument is more naturally stated for the support of E as an operator; the proof should define this notion explicitly or decompose E into Pauli components.
  2. [Lemma 2 proof] The lower bound a n ≤ |S_t| is not used anywhere in the proof of Lemma 2; only the upper bound |S_t| ≤ b n and bounded participation are used. This is harmless but worth clarifying, especially since the unused lower bound is part of the hypothesis set that leads to the vacuity identified above.

Circularity Check

2 steps flagged

Theorem 1's proof assumes the per-triple logical-CCZ mechanism it is supposed to prove; its hypotheses are also mutually inconsistent, so the advertised fountain is vacuous.

specific steps
  1. self definitional [Remark 1 and Appendix E, Step 2]
    "The orthogonality conditions in Definition 1 and Remark 1 ensure that there exists a diagonal unitary U_t built from physical Z, CZ, and CCZ gates whose action on the subspace spanned by these three logical qubits is (up to Clifford conjugation) a logical CCZ gate. Concretely, one may take a transversal CCZ on physical representatives x_t, y_t, z_t, possibly supplemented by local Clifford layers to correct unwanted phases; the exact form is not important here, only the locality pattern."

    The theorem's conclusion is precisely that a constant-depth physical CCZ circuit implements logical CCZ gates on many triples. But Step 2 assumes, rather than derives, that each magic-friendly triple already admits a bounded-locality U_t implementing logical CCZ. Definition 1 states only pairwise orthogonality and odd triple overlap; it does not construct U_t. Remark 1 merely asserts the implication without proof. The paper's own Example 1 shows the three logical supports can overlap at a single qubit, so a 'transversal CCZ on physical representatives' is not a well-defined operation on one code block: the same physical qubit would have to participate in multiple CCZ gates. Thus the Ω(n^γ) logical-CCZ fountain is the assumed per-triple mechanism scaled up by packing and coloring, not a con

  2. other [Theorem 1, hypotheses 1–3]
    "1. (Many candidates) |S_n| ≥ c1 n^{1+γ} ... 2. (Support size) There exist constants 0 < a ≤ b such that an ≤ |S_t| ≤ bn for all t ∈ S_n. 3. (Bounded participation) There is M > 0 such that each coordinate i ∈ [n] belongs to at most M sets S_t with t ∈ S_n."

    The hypotheses are mutually inconsistent. Each |S_t| ≥ a n, so Σ_t |S_t| ≥ |S_n| a n. But condition 3 gives Σ_t |S_t| = Σ_{i=1}^n |{t : i ∈ S_t}| ≤ M n. Hence |S_n| ≤ M/a, an O(1) bound, contradicting condition 1's demand that |S_n| ≥ c_1 n^{1+γ} with γ > 0 for all sufficiently large n. Consequently no CSS code family can satisfy the theorem's hypotheses; the conditional is true only vacuously, and the advertised constant-depth CCZ fountain is not actually derived for any existing family. This is a fatal consistency gap in the derivation chain.

full rationale

The main circularity is localized in Appendix E, Step 2. The theorem's hypotheses count magic-friendly triples defined by pairwise orthogonality and odd triple overlap, but the proof must show that each triple yields a bounded-locality diagonal unitary implementing a logical CCZ. Instead, Step 2 asserts this ('the exact form is not important here'), citing Remark 1, which itself is an unproved sufficiency claim. No construction is given, and Example 1 shows that a 'transversal CCZ on physical representatives' is not even well defined when the three logical supports overlap. The packing lemma and edge-coloring lemma only amplify this assumed per-triple mechanism; the Ω(n^γ) logical-CCZ result is therefore imported rather than derived. Separately, the theorem's hypotheses are unsatisfiable: conditions 2 and 3 imply |S_n| ≤ M/a, while condition 1 requires |S_n| ≥ c_1 n^{1+γ}, so no large-n family can satisfy all three. The theorem is thus vacuous and does not establish a native CCZ magic-state fountain for any code family. This is a correctness failure rather than a strict circularity, but it reinforces that the central claim lacks an independent derivation. There are no significant self-citation or ansatz-smuggling circularities in the paper; the flaws are the assumed per-triple mechanism and the inconsistent hypothesis set.

Axiom & Free-Parameter Ledger

0 free parameters · 3 axioms · 0 invented entities

The paper introduces no fitted parameters and no new physical entities. Its key mathematical assumption is the existence of bounded-locality logical CCZ implementations for magic-friendly triples, which is unproved, plus a triple-counting assumption that is internally inconsistent with the stated support-size and participation bounds.

axioms (3)
  • ad hoc to paper A magic-friendly triple of logical X operators suffices to build a logical CCZ with bounded-locality physical CCZ gates.
    Asserted in Appendix E, Step 2 without proof; this is the central mechanism that would make the theorem work.
  • domain assumption A CSS code family can simultaneously have Omega(n^{1+gamma}) magic-friendly triples, each support of size Theta(n), and per-qubit participation bounded by M.
    Theorem 1, conditions 1-3; this combination is unsatisfiable by double counting for any fixed M and a>0, so the theorem has no instances.
  • standard math Standard stabilizer/CSS formalism and the light-cone distance bound (Lemma 4) correctly map physical to logical operators and preserve distance.
    Background results used throughout; the light-cone bound is standard for constant-depth local circuits.

pith-pipeline@v1.3.0-alltime-deepseek · 9006 in / 17585 out tokens · 195640 ms · 2026-08-04T06:13:32.223742+00:00 · methodology

0 comments
read the original abstract

Native constant-depth non-Clifford gates on quantum low-density parity-check (qLDPC) codes can substantially reduce the space-time overhead of magic-state distillation. This paper investigates the underlying parity and structural conditions governing parallel non-Clifford gate implementations. By analyzing their invariance, algebraic forms, and circuit synthesis, we establish fundamental limits on code scaling: strict saturated implementations cannot achieve constant-depth realization as code distance grows due to tight distance-depth bounds. We evaluate alternative scaling routes, showing that scalable designs must either relax strict subspace requirements or manage gate congestion. Through analytical bounds and code searches, we demonstrate that reconciling parallel non-Clifford operations with linear distance requires navigating these fundamental structural trade-offs.

Figures

Figures reproduced from arXiv: 2601.22489 by Mohammad Rowshan.

Figure 1
Figure 1. Figure 1: FIG. 1. Toy 3-uniform hypergraph with three hyperedges [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Supports of a magic-friendly triple [PITH_FULL_IMAGE:figures/full_fig_p003_3.png] view at source ↗

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

41 extracted references · 20 linked inside Pith

  1. [1]

    (Support size)an≤ |S t| ≤bnfor allt∈ S, for constants0< a≤b

  2. [2]

    A triple(x, y, z)∈(C⊥ Z )3 ismagic- friendlyif: 1.x, y, zrepresent three independent logicalXoper- ators, i.e., their images inLX are linearly indepen- dent

  3. [3]

    pairwise orthogonality holds: ⟨x, y⟩=⟨x, z⟩=⟨y, z⟩= 0 (mod 2);(5)

  4. [4]

    the triple overlap is odd: τ(x, y, z) = 1 (mod 2).(6) 1 2 3 4 5 6 FIG. 1. Toy 3-uniform hypergraph with three hyperedges (triangles), distinguished by three line styles (solid, dashed, dotted). Gates in each style class can be executed in one layer of disjointCCZgates. Remark 1(LogicalZcorrections).In more general CSS constructions, one may allow⟨x (a), x...

  5. [5]

    Then there exists a subcollectionT ⊆ Ssuch that: •each coordinatei∈[n]belongs to at most one sup- portS t witht∈ T; •|T | ≥ |S|/(M bn)

    (Bounded participation) For alli∈[n],|{t∈ S: i∈S t}| ≤M. Then there exists a subcollectionT ⊆ Ssuch that: •each coordinatei∈[n]belongs to at most one sup- portS t witht∈ T; •|T | ≥ |S|/(M bn). A proof is given in Appendix C. Figure 2 gives a schematic view of this packing process. Main structural theorem.We now combine the hyper- graph and packing viewpoi...

  6. [6]

    1 2 3 4 x y z FIG

    (Many candidates)|S n| ≥c1n1+γ for some con- stantsc 1 >0,γ >0. 1 2 3 4 x y z FIG. 3. Supports of a magic-friendly triple(x, y, z)on four qubits forx= (0,1,1,1),y= (1,0,1,1),z= (1,1,0,1). Solid, dashed, and dotted rings mark the qubits in the supports of x,y, andz, respectively. Qubit 4 lies in all three supports and is the unique triple overlap

  7. [7]

    (Support size) There exist constants0< a≤bsuch thatan≤ |St| ≤bnfor allt∈ Sn

  8. [8]

    (Bounded participation) There isM >0such that each coordinatei∈[n]belongs to at mostMsets St witht∈ Sn. Then there exist constantsβ >0,∆,c >0, and, for each n, a collection of triplesT n ⊆ Sn with|T n| ≥cnβ and a bounded-degree 3-uniform hypergraph of physicalCCZ gates of maximum degree∆such that: •there is a constant-depth circuit of physicalCCZ gates th...

  9. [9]

    It isolates a precise combinatorial property of the logicalXspace that any qLDPC-based magic-state fountain must exploit, namely the existence of a large, low-overlap packing of magic-friendly triples. This unifies and generalizes earlier constructions based on triorthogonal matrices and cup-product structures [9, 21, 25] into a single algebraic condi- ti...

  10. [10]

    This suggests a new interface between quantum coding theory and the expanding toolbox of expander and Tanner- code techniques [14, 15]

    It recasts the search for native non-Clifford gates on qLDPC codes as an explicit problem in the combi- natorics of high-dimensional expanders: prove that a given asymptotically good qLDPC family admits Ω(k)magic-friendlytripleswithboundedparticipa- tionM=O(1), and Theorem 1 immediately yields an asymptotically good magic-state fountain with LDPC checks a...

  11. [11]

    Weexpectthesamepatterntoapplyto more general multi-controlled-Zmagic-state facto- ries and to other families of product-based qLDPC codes

    It provides a reusable blueprint for turning struc- tural information about logical operators into con- crete bounded-degree non-Clifford circuits: the combination of packing (Lemma 2), hypergraph edge-coloring (Lemma 1), and the distance- preservation lemma (Lemma 4) yields a general recipe for designing constant-depth diagonal cir- cuits (here,CCZlayers...

  12. [12]

    mark all coordinates inSt as used

  13. [13]

    We stop when no triple with support contained in the unused coordinates remains

    remove fromSevery triplet ′ ∈ Sfor whichSt′ ∩ St ̸=∅. We stop when no triple with support contained in the unused coordinates remains. At that point, no two se- lected supports intersect, and each coordinate belongs to at most oneSt witht∈ T. Bounding removed triples per selection.Fix a triplet chosen during the process. For each coordinatei∈S t, at mostM...

  14. [14]

    Aharonov and M

    D. Aharonov and M. Ben-Or, Fault-tolerant quantum computation with constant error, inProceedings of the 29th Annual ACM Symposium on Theory of Computing (STOC)(1997) pp. 176–188, arXiv:quant-ph/9611025

    Pith/arXiv arXiv 1997

  15. [15]

    A. Y. Kitaev, Quantum computations: algorithms and error correction, Russian Mathematical Surveys52, 1191 (1997)

    1997

  16. [16]

    Knill, R

    E. Knill, R. Laflamme, and W. H. Zurek, Resilient quan- tumcomputation: Errormodelsandthresholds,Proceed- ings of the Royal Society A454, 365 (1998)

    1998

  17. [17]

    Aliferis, D

    P. Aliferis, D. Gottesman, and J. Preskill, Quantum accuracy threshold for concatenated distance-3 codes, Quantum Information and Computation6, 97 (2006), arXiv:quant-ph/0504218

    Pith/arXiv arXiv 2006

  18. [18]

    A. Y. Kitaev, Fault-tolerant quantum computation by anyons, Annals of Physics303, 2 (2003), arXiv:quant- ph/9707021

    arXiv 2003

  19. [19]

    Bombin and M

    H. Bombin and M. A. Martin-Delgado, Topological quan- tum distillation, Physical Review Letters97, 180501 (2006), arXiv:quant-ph/0605138

    Pith/arXiv arXiv 2006

  20. [20]

    Kubica and M

    A. Kubica and M. E. Beverland, Universal transversal gates with color codes: A simplified approach, Physical Review A91, 032330 (2015), arXiv:1410.0069

    Pith/arXiv arXiv 2015

  21. [21]

    Bravyi and A

    S. Bravyi and A. Kitaev, Universal quantum computa- tion with ideal clifford gates and noisy ancillas, Physical Review A71, 022316 (2005), arXiv:quant-ph/0403025

    Pith/arXiv arXiv 2005

  22. [22]

    Bravyi and J

    S. Bravyi and J. Haah, Magic-state distillation with low overhead, Physical Review A86, 052329 (2012), arXiv:1209.2426

    Pith/arXiv arXiv 2012

  23. [23]

    E. T. Campbell and J. O’Gorman, Quantum computa- tion with realistic magic-state factories, Physical Review A95, 032338 (2017), arXiv:1606.01904

    Pith/arXiv arXiv 2017

  24. [24]

    Gidney and A

    C. Gidney and A. G. Fowler, Efficient magic state facto- ries with a catalyzed ccz to 2t transformation, Quantum 3, 135 (2019), arXiv:1812.01238. 7

    Pith/arXiv arXiv 2019

  25. [25]

    Y. R. Sanders, D. W. Berry, P. C. S. Costa, L. W. Tessler, N. Wiebe, C. Gidney, H. Neven, and R. Babbush, Com- pilation of fault-tolerant quantum heuristics for combi- natorial optimization, PRX Quantum1, 020312 (2020), arXiv:2007.07391

    Pith/arXiv arXiv 2020

  26. [26]

    Panteleev and G

    P. Panteleev and G. Kalachev, Asymptotically good quantum and locally testable classical LDPC codes, IEEE Transactions on Information Theory70, 4836 (2024), arXiv:2111.03654

    Pith/arXiv arXiv 2024

  27. [27]

    Leverrier and G

    A. Leverrier and G. Zémor, Quantum Tanner codes, in63rd Annual IEEE Symposium on Foundations of Computer Science (FOCS)(2022) pp. 872–883, arXiv:2202.13641

    Pith/arXiv arXiv 2022

  28. [28]

    N. P. Breuckmann and J. N. Eberhardt, Quantum low- density parity-check codes, PRX Quantum2, 040101 (2021), arXiv:2103.06309

    Pith/arXiv arXiv 2021

  29. [29]

    Bravyi, A

    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, Nature627, 778 (2024)

    2024

  30. [30]

    Pecorari, S

    L. Pecorari, S. Jandura, G. K. Brennen, and G. Pupillo, High-rate quantum LDPC codes for long- range-connected neutral atom registers, Nature Commu- nications16, 1 (2025), arXiv:2404.13010

    Pith/arXiv arXiv 2025

  31. [31]

    Q. Xu, H. Zhou, G. Zheng, D. Bluvstein, J. P. B. Ataides, M. D. Lukin, and L. Jiang, Fast and parallelizable logi- cal computation with homological product codes, Phys- ical Review X15, 10.1103/PhysRevX.15.021065 (2025), arXiv:2407.18490

    Pith/arXiv arXiv 2025

  32. [32]

    Baspin and A

    N. Baspin and A. Krishna, Quantifying nonlocality: How outperforminglocalquantumcodesisexpensive,Physical Review Letters129, 050505 (2022), arXiv:2112.05776

    Pith/arXiv arXiv 2022

  33. [33]

    M. A. Tremblay, N. Delfosse, and M. E. Beverland, Constant-overhead quantum error correction with thin planar connectivity, Physical Review Letters129, 050504 (2022), arXiv:2109.14609

    Pith/arXiv arXiv 2022

  34. [34]

    Golowich and V

    L. Golowich and V. Guruswami, Asymptotically good quantum codes with transversal non-clifford gates, arXiv preprint (2024), arXiv:2408.09254

    Pith/arXiv arXiv 2024

  35. [35]

    Golowich and T

    L. Golowich and T. Lin, Quantum LDPC codes with transversal non-clifford gates via products of algebraic codes, arXiv preprint (2024), arXiv:2410.14662 [quant- ph]

    Pith/arXiv arXiv 2024

  36. [36]

    Lin, Transversal non-clifford gates for quan- tum LDPC codes on sheaves, arXiv preprint (2024), arXiv:2410.14631

    T.-C. Lin, Transversal non-clifford gates for quan- tum LDPC codes on sheaves, arXiv preprint (2024), arXiv:2410.14631

    Pith/arXiv arXiv 2024

  37. [37]

    Z. He, V. Vaikuntanathan, A. Wills, and R. Y. Zhang, Asymptotically good quantum codes with addressable and transversal non-clifford gates, arXiv preprint (2025), arXiv:2507.05392 [quant-ph]

    Pith/arXiv arXiv 2025

  38. [38]

    G.Zhu,Atopologicaltheoryforqldpc: Non-cliffordgates and magic state fountain on homological product codes with constant rate and beyond theN1/3 distance barrier, arXiv preprint (2025), arXiv:2501.19375 [quant-ph]

    arXiv 2025

  39. [39]

    J. Haah, M. B. Hastings, D. Poulin, and D. Wecker, Magic state distillation with low space overhead and op- timal asymptotic input count, Quantum1, 31 (2017)

    2017

  40. [40]

    Bravyi and R

    S. Bravyi and R. König, Classification of topologically protected gates for local stabilizer codes, Physical Review Letters110, 1 (2013)

    2013

  41. [41]

    Pastawski and B

    F. Pastawski and B. Yoshida, Fault-tolerant logical gates in quantum error-correcting codes, Physical Review A 91, 1 (2015)

    2015