pith. sign in

arxiv: 2606.22853 · v1 · pith:UIRPR6IQnew · submitted 2026-06-22 · 🪐 quant-ph

Bunny Codes: Broadening Superconducting Quantum Error Correction Capability through Advanced Control Engineering

Runshi Zhou , Xingye Yuan , Linghang Kong , Fang Zhang , Kai Zhang , Zhaohui Yang , Jianxin Chen This is my paper

Pith reviewed 2026-06-26 08:22 UTC · model grok-4.3

classification 🪐 quant-ph
keywords Bunny codesqLDPC codesquantum error correctionsuperconducting qubitsCXSWAP gatetoric codesurface codecode rate
0
0 comments X

The pith

Bunny codes achieve three times the code rate of toric codes on nearest-neighbor superconducting qubit layouts by using an expanded native gate set.

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

The paper establishes that an enlarged set of native two-qubit gates, specifically CNOT together with CXSWAP, allows certain quantum low-density parity-check codes with nonlocal stabilizers to be implemented on two-dimensional nearest-neighbor connectivity without long-range interactions. Exhaustive search identifies a family of such codes, called Bunny codes, whose best weight-6 instances in periodic boundary conditions reach roughly three times the code rate of the toric code and retain a twofold advantage over the rotated surface code after conversion to open boundaries. Circuit-level simulations further show that selected Bunny codes can reach logical error rates an order of magnitude below those of toric codes at comparable rates. A reader would care because the result replaces the engineering burden of long-range couplers with a control-engineering change that is already accessible in superconducting hardware.

Core claim

Drawing on advances in superconducting qubit control that unlock enriched native gate sets, exhaustive search identifies qLDPC codes that admit syndrome extraction using only CNOT and CXSWAP on nearest-neighbor connectivity while matching the performance of direct CNOT implementations that would otherwise require long-range interactions; the resulting Bunny codes with weight-6 stabilizers deliver approximately three times the code rate of the toric code under periodic boundaries and twice the rate of the rotated surface code under open boundaries, with some members exhibiting an order-of-magnitude reduction in logical error rate under circuit-level noise.

What carries the argument

Bunny codes: the qLDPC codes discovered by exhaustive search that admit syndrome extraction circuits using the two-qubit gate pool {CNOT, CXSWAP} on two-dimensional nearest-neighbor connectivity, thereby realizing nonlocal stabilizers without long-range couplers.

If this is right

  • High-rate quantum error correction becomes feasible on existing superconducting hardware without fabrication of long-range couplers.
  • Code-rate advantages of roughly 3× (periodic) and 2× (open) are obtained across examined distances for weight-6 stabilizer Bunny codes.
  • Selected Bunny codes already exhibit lower logical error rates than toric codes of comparable rate under circuit-level noise.
  • Hardware complexity for quantum error correction is reduced by substituting an expanded gate set for additional coupler hardware.

Where Pith is reading between the lines

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

  • The same gate-set expansion might be tested on other qLDPC families or on hardware platforms that already support CXSWAP-like operations.
  • If CXSWAP error rates prove higher than assumed, the logical-error advantage could be recovered by optimizing decoder or circuit scheduling rather than by changing the code itself.
  • Bunny codes could serve as drop-in replacements for surface-code patches in near-term processors, potentially increasing the number of logical qubits per physical qubit area.
  • The search method used to find Bunny codes could be applied to other enriched gate sets that become available through further control advances.

Load-bearing premise

CXSWAP gates can be realized with fidelity comparable to CNOT and without introducing correlated errors that would erase the reported logical-error advantage.

What would settle it

A circuit-level simulation or hardware experiment that includes a realistic error model for CXSWAP gates and shows that the logical error rates of the reported Bunny codes no longer remain an order of magnitude below those of toric codes at matched rates.

Figures

Figures reproduced from arXiv: 2606.22853 by Fang Zhang, Jianxin Chen, Kai Zhang, Linghang Kong, Runshi Zhou, Xingye Yuan, Zhaohui Yang.

Figure 1
Figure 1. Figure 1: (a) Six data qubits (labeled di) and two ancilla qubits (labeled X and Z) in a larger surface code lattice. These qubits support two stabilizers of the surface code, Z1Z2Z3Z4 and X3X4X5X6. (b) Example of a syndrome extraction circuit for measuring these two stabilizers using only CNOT gates. Note that all CNOT gates are between adjacent qubits in the layout. However, there is another direction for implemen… view at source ↗
Figure 2
Figure 2. Figure 2: (a) An example of basic units on a hardware with hexagonal connectiv [PITH_FULL_IMAGE:figures/full_fig_p006_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: An example of a layer of gates generated by an action of [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: An example of the movement of a sub-lattice by a layer of gates [PITH_FULL_IMAGE:figures/full_fig_p009_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: The parallelized syndrome extraction circuit constructed from the [PITH_FULL_IMAGE:figures/full_fig_p011_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: A demonstration of using padding qubits to facilitate routing at the [PITH_FULL_IMAGE:figures/full_fig_p014_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: Diagrams for some qubit connectivities used in this paper. (a) [PITH_FULL_IMAGE:figures/full_fig_p015_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: Highest code rates achieved for quantum error correction codes found [PITH_FULL_IMAGE:figures/full_fig_p017_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The logical error rate per round per logical qubit of the [PITH_FULL_IMAGE:figures/full_fig_p018_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: The logical error rate per round per logical qubit for some representa [PITH_FULL_IMAGE:figures/full_fig_p018_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Highest code rates achieved for quantum error correction codes with [PITH_FULL_IMAGE:figures/full_fig_p020_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: The logical error rate per round per logical qubit for some repre [PITH_FULL_IMAGE:figures/full_fig_p020_12.png] view at source ↗
read the original abstract

Drawing on advances in superconducting qubit control schemes that unlock enriched native gate sets at the hardware level, we systematically examine how harnessing this enlarged physical two-qubit gate pool -- specifically CNOT and CXSWAP -- streamlines syndrome extraction for certain qLDPC codes with nonlocal stabilizers. Through an exhaustive search, we discover a set of qLDPC codes with various stabilizer weights and distances that can be implemented on the two-dimensional nearest-neighbor qubit connectivity native to superconducting hardware while achieving performance equivalent to that of the direct CNOT implementation requiring long-range interactions. We refer to those codes as Bunny codes. Across all code distances we examine, the best Bunny codes with weight-6 stabilizers in periodic boundary conditions have a code rate approximately $3\times$ that of the toric code; when converted to open boundary conditions, they retain an approximately $2\times$ code rate advantage over the rotated surface code. In circuit-level simulation, we find that some Bunny codes exhibit logical error rates an order of magnitude lower than toric codes with comparable code rates. Our results demonstrate that high-performance quantum error correction can be achieved using an expanded gate set rather than long-range couplers, thereby significantly reducing hardware complexity.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

2 major / 2 minor

Summary. The paper introduces Bunny codes, a family of qLDPC codes discovered via exhaustive search that exploit an expanded native gate set (CNOT + CXSWAP) on 2D nearest-neighbor superconducting qubit connectivity. The central claims are that the best weight-6 Bunny codes achieve ~3× the code rate of the toric code under periodic boundaries and ~2× that of the rotated surface code under open boundaries, while selected instances exhibit an order-of-magnitude lower logical error rate than toric codes of comparable rate in circuit-level simulations.

Significance. If the reported code-rate and logical-error advantages survive realistic hardware noise, the approach would meaningfully reduce the need for long-range couplers in superconducting QEC, lowering hardware complexity while retaining high performance. The use of an enlarged gate set to enable better qLDPC codes on planar lattices is a concrete engineering contribution.

major comments (2)
  1. [Circuit-level simulations] Circuit-level simulation section: the headline claim of order-of-magnitude lower logical error rates rests on the assumption that CXSWAP can be treated as fidelity-equivalent to CNOT with no additional correlated errors or crosstalk; the manuscript provides no hardware error model, calibration data, or pulse-sequence analysis to support this equivalence, which is load-bearing for the performance comparison.
  2. [Exhaustive search methodology] Exhaustive search and code construction section: the search criteria, distance and rate definitions, stabilizer-weight constraints, and exact parameters of the discovered Bunny codes are not specified in sufficient detail to reproduce the reported 3×/2× rate advantages or to verify that the codes remain implementable under the stated connectivity.
minor comments (2)
  1. [Introduction and results] Notation for code parameters (n,k,d) and boundary conditions should be stated explicitly when comparing Bunny codes to toric and surface codes.
  2. [Figures] Figure captions for simulation plots should include the precise noise model parameters and number of shots used.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback. We address each major comment below with clarifications and indicate where the manuscript will be revised for improved clarity and reproducibility.

read point-by-point responses

  1. Referee: Circuit-level simulation section: the headline claim of order-of-magnitude lower logical error rates rests on the assumption that CXSWAP can be treated as fidelity-equivalent to CNOT with no additional correlated errors or crosstalk; the manuscript provides no hardware error model, calibration data, or pulse-sequence analysis to support this equivalence, which is load-bearing for the performance comparison.

    Authors: We acknowledge the validity of this point: our circuit-level simulations model CXSWAP as having equivalent error rates to CNOT, based on the premise that advanced control engineering enables the enriched gate set without introducing new dominant error channels. The manuscript does not include hardware-specific error models or calibration data, as the work is a theoretical exploration of code performance under this gate set. We will revise the simulation section to explicitly state this modeling assumption, discuss the potential impact of unmodeled effects such as crosstalk, and qualify the results as indicative of the approach's promise pending experimental validation. revision: yes

  2. Referee: Exhaustive search and code construction section: the search criteria, distance and rate definitions, stabilizer-weight constraints, and exact parameters of the discovered Bunny codes are not specified in sufficient detail to reproduce the reported 3×/2× rate advantages or to verify that the codes remain implementable under the stated connectivity.

    Authors: We agree that the current description of the exhaustive search lacks sufficient detail for full reproducibility. The revised manuscript will expand the code construction section to specify the search algorithm parameters, precise definitions of code distance and rate, stabilizer weight constraints used, and the explicit parameters (including stabilizer supports and qubit connectivity graphs) for the reported Bunny codes that achieve the stated rate advantages. revision: yes

Circularity Check

0 steps flagged

No significant circularity; claims rest on search and simulation outputs

full rationale

The paper's central claims (higher code rates for Bunny codes vs toric/rotated surface codes, and lower logical error rates in circuit simulations) are presented as results of an exhaustive search over qLDPC codes using CNOT+CXSWAP gates on nearest-neighbor connectivity, followed by circuit-level simulations. No equations, definitions, or performance metrics are shown to reduce to fitted parameters, self-referential quantities, or load-bearing self-citations. The abstract and provided text frame the code-rate advantages and error-rate comparisons as direct outputs of the search and simulation pipeline rather than by-construction renamings or ansatzes imported from prior author work. The CXSWAP fidelity assumption is a modeling choice whose validity is external to the derivation itself and does not create circularity within the reported results.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review provides no identifiable free parameters, axioms, or invented entities beyond standard quantum error correction assumptions such as independent Pauli noise models in simulation.

pith-pipeline@v0.9.1-grok · 5758 in / 1197 out tokens · 26997 ms · 2026-06-26T08:22:01.367573+00:00 · methodology

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

36 extracted references · 6 canonical work pages

  1. [1]

    Shor, P. W. Polynomial-Time Algorithms for Prime Factorization and Dis- crete Logarithms on a Quantum Computer.SIAM Journal on Computing 26, 1484–1509 (1997). Place: USA Publisher: Society for Industrial and Applied Mathematics

    1997

  2. [2]

    Grover, L. K. A fast quantum mechanical algorithm for database search. InProceedings of the Twenty-Eighth Annual ACM Symposium on Theory of Computing, STOC ’96, 212–219 (Association for Computing Machinery, New York, NY, USA, 1996). Event-place: Philadelphia, Pennsylvania, USA

    1996

  3. [3]

    G., Mariantoni, M., Martinis, J

    Fowler, A. G., Mariantoni, M., Martinis, J. M. & Cleland, A. N. Surface codes: Towards practical large-scale quantum computation.Physical Re- view A86, 032324 (2012). URLhttps://link.aps.org/doi/10.1103/ PhysRevA.86.032324

    2012

  4. [4]

    23 Table 6: The index of all the action on the square grid connectivity

    Acharya, R.et al.Quantum error correction below the surface code thresh- old.Nature638, 920 (2024). 23 Table 6: The index of all the action on the square grid connectivity. Index Coupler Gate 0 (0,0)↔(1,0) CNOT 1 (0,0)↔(1,0) CXSWAP 2 (0,0)↔(−1,0) CNOT 3 (0,0)↔(−1,0) CXSWAP 4 (0,0)↔(0,1) CNOT 5 (0,0)↔(0,1) CXSWAP 6 (0,0)↔(0,−1) CNOT 7 (0,0)↔(0,−1) CXSWAP 8...

    2024

  5. [5]

    & Zemor, G

    Tillich, J.-P. & Zemor, G. Quantum LDPC codes with positive rate and minimum distance proportional to the square root of the blocklength.IEEE Transactions on Information Theory60, 11931202 (2014). URLhttp: //dx.doi.org/10.1109/TIT.2013.2292061

    work page doi:10.1109/tit.2013.2292061 2014

  6. [6]

    URLhttps://doi.org/10.1038/ s41586-024-07107-7

    Bravyi, S.et al.High-threshold and low-overhead fault-tolerant quantum memory.Nature627, 778–782 (2024). URLhttps://doi.org/10.1038/ s41586-024-07107-7

    2024

  7. [7]

    URL https://doi.org/10.1145/3779066

    Viszlai, J.et al.qSIEVE: Efficient qldpc memory via systolic movement in atom arrays.ACM Transactions on Quantum Computing7(2026). URL https://doi.org/10.1145/3779066

    work page doi:10.1145/3779066 2026

  8. [8]

    & Pryadko, L

    Wang, R., Lin, H.-K. & Pryadko, L. P. Abelian and non-abelian quantum two-block codes.arXiv preprint arXiv:2305.06890(2023). URLhttps: //arxiv.org/abs/2305.06890.2305.06890

    arXiv 2023

  9. [9]

    URLhttps://doi.org/10

    Wang, K.et al.Demonstration of low-overhead quantum error correction codes.Nature Physics22, 308–314 (2026). URLhttps://doi.org/10. 1038/s41567-025-03157-4. Publisher: Nature Publishing Group

    2026

  10. [10]

    Mathews, M.et al.Placing and routing non-local quantum error correct- ing codes in multi-layer superconducting qubit hardware.arXiv preprint arXiv:2507.23011(2025).2507.23011

    Pith/arXiv arXiv 2025

  11. [11]

    Chen, J., Ding, D., Gong, W., Huang, C. & Ye, Q. One Gate Scheme to Rule Them All: Introducing a Complex Yet Reduced Instruction Set for 24 Quantum Computing. InProceedings of the 29th ACM International Con- ference on Architectural Support for Programming Languages and Operat- ing Systems, Volume 2, ASPLOS ’24, 779–796 (Association for Computing Machiner...

    arXiv 2024

  12. [12]

    URLhttps: //www.nature.com/articles/s41567-025-02990-x

    Chen, Z.et al.Efficient implementation of arbitrary two-qubit gates us- ing unified control.Nature Physics21, 1489–1496 (2025). URLhttps: //www.nature.com/articles/s41567-025-02990-x. Publisher: Nature Publishing Group

    2025

  13. [13]

    Under prepara- tion

    Wang, H.et al.Lifting connectivity bottlenecks in superconducting quan- tum processors via enriched native two-qubit gates (2026). Under prepara- tion

    2026

  14. [14]

    Relaxing hardware requirements for surface code circuits using time- dynamics.Quantum, 7:1172, Nov 2023

    McEwen, M., Bacon, D. & Gidney, C. Relaxing Hardware Requirements for Surface Code Circuits using Time-dynamics.Quantum7, 1172 (2023). URLhttps://doi.org/10.22331/q-2023-11-07-1172

    work page doi:10.22331/q-2023-11-07-1172 2023

  15. [15]

    URLhttps://arxiv.org/abs/2412.14360

    Eickbusch, A.et al.Demonstrating dynamic surface codes.arXiv preprint arXiv:2412.14360(2025). URLhttps://arxiv.org/abs/2412.14360. 2412.14360

    arXiv 2025

  16. [16]

    P., Byfield, D

    Gehr, G. P., Byfield, D. & Ruban, A. Directional codes: a new family of quantum LDPC codes on hexagonal- and square-grid connectivity hard- ware.arXiv preprint arXiv:2507.19430(2025).2507.19430

    arXiv 2025

  17. [17]

    To appear in ISCA ’26,2511.04608

    Yang, Z.et al.Qubit mapping and routing tailored to advanced quantum isas: Not as costly as you think.arXiv preprint arXiv:2511.04608(2025). To appear in ISCA ’26,2511.04608

    Pith/arXiv arXiv 2025

  18. [18]

    Zhou, R.et al.Louvre: Relaxing hardware requirements of quantum LDPC codes by routing with expanded quantum instruction set.arXiv preprint arXiv:2508.20858(2025).2508.20858

    arXiv 2025

  19. [19]

    Liang, Z., Eberhardt, J. N. & Chen, Y.-A. Planar quantum low-density parity-check codes with open boundaries.arXiv preprint arXiv:2504.08887 (2025).2504.08887

    arXiv 2025

  20. [20]

    URLhttps://arxiv.org/abs/2606.19482

    Gu, B.et al.Nearest-neighbour gates are all you need: High-rate quan- tum low-density parity-check codes on a planar grid.arXiv preprint arXiv:2606.19482(2026). URLhttps://arxiv.org/abs/2606.19482. 2606.19482

    Pith/arXiv arXiv 2026

  21. [21]

    M., McLauchlan, C

    Nixon, G. M., McLauchlan, C. K. & van Rest, C. C. L. Vine codes: Low- overhead quantum ldpc codes on a planar square grid.arXiv preprint arXiv:2606.20263(2026). URLhttps://arxiv.org/abs/2606.20263. 2606.20263. 25

    Pith/arXiv arXiv 2026

  22. [22]

    Yang, Z.et al.Reconfigurable quantum instruction set computers for high performance attainable on hardware. InProceedings of the 31st ACM International Conference on Architectural Support for Programming Lan- guages and Operating Systems, Volume 2, ASPLOS ’26, 15231546 (As- sociation for Computing Machinery, New York, NY, USA, 2026). URL https://doi.org/1...

    work page doi:10.1145/3779212.3790208 2026

  23. [23]

    & Cong, J

    Tan, B. & Cong, J. Optimal qubit mapping with simultaneous gate absorp- tion. In2021 IEEE/ACM International Conference On Computer Aided Design (ICCAD), 1–8 (IEEE, 2021)

    2021

  24. [24]

    & McEwen, M

    Gidney, C., Newman, M. & McEwen, M. Benchmarking the Planar Honey- comb Code.Quantum6, 813 (2022). URLhttps://doi.org/10.22331/ q-2022-09-21-813

    2022

  25. [25]

    P., Crawford, O

    Geh´ er, G. P., Crawford, O. & Campbell, E. T. Tangling schedules eases hardware connectivity requirements for quantum error correction.PRX Quantum5, 010348 (2024). URLhttps://link.aps.org/doi/10.1103/ PRXQuantum.5.010348

    2024

  26. [26]

    In2026 Design, Automation & Test in Europe Conference (DATE), 1–7 (IEEE, 2026)

    Zhang, K.et al.Optimal compilation of syndrome extraction circuits for general quantum LDPC codes. In2026 Design, Automation & Test in Europe Conference (DATE), 1–7 (IEEE, 2026)

    2026

  27. [27]

    Shaw, M. H. & Terhal, B. M. Optimising quantum error correction using morphing circuits.arXiv preprint arXiv:2604.09797(2026).2604.09797

    Pith/arXiv arXiv 2026

  28. [28]

    Kovalev, A. A. & Pryadko, L. P. Improved quantum hypergraph-product LDPC codes. In2012 IEEE International Symposium on Information Theory Proceedings, 348352 (IEEE, 2012). URLhttp://dx.doi.org/10. 1109/ISIT.2012.6284206

    arXiv 2012

  29. [29]

    PRX Quantum , volume =

    Liang, Z., Liu, K., Song, H. & Chen, Y.-A. Generalized toric codes on twisted tori for quantum error correction.PRX Quantum6, 020357 (2025). URLhttps://link.aps.org/doi/10.1103/rmy6-9n89

    work page doi:10.1103/rmy6-9n89 2025

  30. [30]

    J., Anderson, J

    Landahl, A. J., Anderson, J. T. & Rice, P. R. Fault-tolerant quantum computing with color codes.arXiv preprint arXiv:1108.5738(2011).1108. 5738

    Pith/arXiv arXiv 2011

  31. [31]

    Pecos: Performance estimator of codes on surfaces

    Ryan-Anderson, C. Pecos: Performance estimator of codes on surfaces. https://github.com/PECOS-packages/PECOS(2018)

    2018

  32. [32]

    & Marques-Silva, J

    Ignatiev, A., Morgado, A. & Marques-Silva, J. PySAT: A Python toolkit for prototyping with SAT oracles. InSAT, 428–437 (2018). URLhttps: //doi.org/10.1007/978-3-319-94144-8_26

    work page doi:10.1007/978-3-319-94144-8_26 2018

  33. [33]

    Ignatiev, A., Tan, Z. L. & Karamanos, C. Towards universally accessible SAT technology. InSAT, 4:1–4:11 (2024). URLhttps://doi.org/10. 4230/LIPIcs.SAT.2024.16. 26

    2024

  34. [34]

    Stim: a fast stabilizer circuit simulator.Quantum5, 497 (2021)

    Gidney, C. Stim: a fast stabilizer circuit simulator.Quantum5, 497 (2021). Publisher: Verein zur Forderung des Open Access Publizierens in den Quantenwissenschaften

    2021

  35. [35]

    2506.01779

    Mller, T.et al.Improved belief propagation is sufficient for real-time decoding of quantum memory.arXiv preprint arXiv:2506.01779(2025). 2506.01779

    arXiv 2025

  36. [36]

    N., Pereira, F

    Eberhardt, J. N., Pereira, F. R. F. & Steffan, V. Pruning qLDPC codes: Towards bivariate bicycle codes with open boundary conditions.arXiv preprint arXiv:2412.04181(2024).2412.04181. 27

    arXiv 2024