Pith. sign in

REVIEW 4 major objections 5 minor 47 references

Multimode rotationally symmetric bosonic codes from group-theoretic construction

T0 review · 4 major / 5 minor · reviewed 2026-08-15 · deepseek-v4-flash

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

Pith's one-line read The paper introduces multimode rotationally symmetric bosonic codes in which, up to the optimal rotational order, dephasing protection improves without sacrificing photon-loss protection, and which correct correlated dephasing exactly.

desk verdict Solid group-theoretic code family with a clean correlated-dephasing result, but the analytic dephasing recovery needs re-derivation and the numerics need release before the no-trade-off claim is fully supported. read the letter →

arxiv 2508.20647 v3 pith:LABI5DML submitted 2025-08-28 quant-ph

Rabsan Galib Ahmed , Adithi Udupa , Giulia Ferrini This is my paper · ORCID
classification quant-ph
keywords bosoniccodesrotationalsymmetryquantumerrorcorrectionmultimodebinomiallinearopticsdephasingnoisequdits
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

The paper aims to establish that rotationally symmetric bosonic codes do not have to choose between protecting against photon loss and protecting against dephasing. It constructs a family of multimode codes by a group-theoretic prescription that starts from the logical Pauli gates one wants to implement with linear optics, and shows that in the two-mode binomial instances the dephasing performance improves as the rotational order $N$ grows while the loss performance stays comparable to the single-mode code. It further claims that the entire two-mode family corrects correlated dephasing noise exactly. A sympathetic reader would care because this removes a known design trade-off and supplies bosonic qudits whose Pauli gates are simple passive-linear operations.

What carries the argument

The machinery is a group-theoretic code construction: choose the group $G=\langle g,h\,|\,h^{2N}=e\rangle$, let the logical representation send $g$ to $X$ and $h$ to $Z$, and build the physical representation from passive linear optics—a beam-splitter network $\hat{U}_{\mathrm{BS}}$ followed by rotations such as $\pi(h)=\hat{U}_{\mathrm{BS}}\exp(i\frac{2\pi}{Nd}\hat{a}_1^\dagger\hat{a}_1)\hat{U}_{\mathrm{BS}}^\dagger$. The new element is the mode-mixing unitary $\hat{U}_{\mathrm{BS}}$, which rotates the noise operators into a basis where dephasing acts partly through generators $\hat{G}^{\pm}$ rather than only through individual photon numbers. The codewords are superpositions of Fock states in which each mode has only levels at multiples of $N$, cyclically shifted across the $d$ modes, so the full Pauli group is realized by beam splitters and rotations. Recovery uses modular number measurements followed by correction unitaries; the exact correlated-dephasing correction uses controlled-$X$ gates of the form $\exp(i\frac{\pi}{2N}\hat{n}_1\otimes\hat{G}^-_{34})$.

What would settle it

Compute the optimal-recovery entanglement infidelity for the $N=4$, $K=2$ code with the optimal angles under pure dephasing with unequal mode rates $\gamma_1\neq\gamma_2$; if the infidelity no longer improves over the single-mode code as $N$ grows, the no-trade-off claim is tied to the equal-rate local noise model. For the $N=2$ case, a closed-form recovery cannot exist at first order because of the Knill-Laflamme violation in Appendix D, so any claimed dephasing advantage should be checked against the numerical semi-definite program at stronger dephasing strengths.

Watch

Extended reading notes

Core claim

The central claim is that adding a second mode, together with a tunable beam-splitter rotation, converts the single-mode rotationally symmetric binomial code into a code with simultaneous resistance to loss and dephasing. Concretely, for the two-mode $K=2$ code with codewords $|0_N\rangle=\hat{U}_{\mathrm{BS}}\frac{1}{\sqrt{2}}(|0\rangle+|2N\rangle)\otimes|N\rangle$ and $|1_N\rangle=\hat{U}_{\mathrm{BS}}|N\rangle\otimes\frac{1}{\sqrt{2}}(|0\rangle+|2N\rangle)$, choosing the beam-splitter angles $\delta=\pi/4$ or $3\pi/4$ and $\phi=\pi/(2N)$ makes the Knill-Laflamme conditions hold to first order for both loss and dephasing when $N=4$, while for $N=2$ only loss admits the closed-form first-order recovery and dephasing is handled by numerically optimized recovery. The paper also proves that any code of the form in Eqs. (4)-(5) corrects arbitrary correlated dephasing of the form in Eq. (11) exactly, using the four-mode circuit in Fig. 3, and that the same construction encodes qudits of any dimension $d$ in $d$ modes. In the paper's own terms, this means the single-mode trade-off between dephasing and loss protection is resolved, up to the optimal value of $N$.

Load-bearing premise

The load-bearing premise is that the noise is well described by independent, Markovian, equal-rate loss and dephasing in each mode, and that correcting to first order in the noise strengths is enough to establish the performance advantage.

Editorial extensions

If this is right

  • For the two-mode $K=2$ binomial codes at optimal angles, dephasing infidelity decreases as $N$ increases up to an optimal value, while loss performance remains comparable to the corresponding single-mode code; the single-mode trade-off is absent in this regime.
  • The full Pauli group on the encoded qubit or qudit is implemented by passive linear optics, so logical $X$ and $Z$ are available without auxiliary qubits.
  • Any two-mode RSB code in the family corrects arbitrary correlated dephasing of the form in Eq. (11) exactly, with a fixed four-mode circuit independent of the order $N$ and the encoding angles.
  • The construction supports qudit encoding in arbitrary dimension $d$, with order-$N$ rotational symmetry in each of $d$ modes.
  • The Hadamard, $S$, $T$, and $CZ$ gates can be implemented with Kerr interactions and gate teleportation, so the code family is compatible with a useful non-universal gate set.

Reading between the lines

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

  • Because the advantage is demonstrated for independent equal-rate local noise, an immediate test is asymmetric rates: if the dephasing improvement persists when $\gamma_1\neq\gamma_2$ or $\kappa_1\neq\kappa_2$, the construction is more robust than the paper's stated model.
  • The exact correlated-dephasing argument relies on commutation of the total photon number with the mode-mixing generators; the same commutation suggests the circuit may extend to correlated loss or to more than two modes, though the paper does not claim this.
  • The torus phase-distribution picture implies the beam-splitter angles are a tunable resource: optimizing them for a specific noise channel could yield further gains, and the same idea might transfer to translationally symmetric bosonic codes.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 5 minor

Summary. The paper introduces a family of multimode rotationally symmetric bosonic codes constructed by inverting the group-theoretic framework of Denys and Leverrier: one first chooses logical Pauli gates implemented by passive linear optics and then derives codewords. The main instances studied are two-mode binomial codes with codewords given by Eqs. (6)-(7), for which the paper claims improved dephasing protection and, unlike single-mode RSB codes, no trade-off between dephasing and loss protection. The paper also claims exact correction of correlated dephasing via a four-mode circuit, and qudit encoding in arbitrary dimensions. The analytical results include first-order recovery maps for loss (N=2,4) and dephasing (N=4), numerical SDP benchmarks showing dephasing infidelity decreasing with N, and an appendix proving the correlated-dephasing circuit.

Significance. If the claims are valid, the code family is a meaningful step for bosonic QEC: it preserves linear-optics logical Pauli gates while improving dephasing resilience, and the exact correction of correlated dephasing is a clean, testable result. The group-theoretic inversion is a promising design principle, and the two-mode binomial instances are explicit enough for experimental follow-up. The analytical structure (Appendices A-D) and the correlated-dephasing proof (Appendix G) are the main strengths. However, several load-bearing analytical statements are internally inconsistent or insufficiently specified, and the qudit generalization appears to have a group-representation flaw. These issues need to be resolved before the central claims can be accepted.

major comments (4)
  1. [Main text, paragraph after Eq. (7); Appendix D, Eqs. (D1)-(D26)] The claim that the N=4 two-mode code satisfies the Knill-Laflamme conditions for dephasing up to first order is inconsistent with the recovery map actually presented. If the KL conditions held for the full set of first-order error operators, the standard recovery would consist of a projective syndrome measurement followed by a single unitary per syndrome; the additional projection P_L/P_E introduced in Eqs. (D17)-(D22) would be unnecessary. Its presence indicates that the (0,0)-syndrome error operators are not proportional on the code, so the KL conditions are not satisfied as stated. The authors should either prove the KL conditions for the complete error set or revise the main-text claim to state that the errors are correctable via a syndrome measurement augmented by a second-level measurement.
  2. [Appendix D, Eqs. (D1)-(D4)] The operators M0^(1) and M0^(2) are defined with scaling γ_i t, e.g. M0^(1)=γ1 t(a1†a1 cos²δ+a2†a2 sin²δ). With this scaling, the terms M0^(1) ρ M0^(1)† in Eq. (D1) are of order (γt)², whereas the first-order dephasing jump terms in Eq. (B14) are of order γt. The decomposition in Eq. (D1) therefore does not match the first-order channel expansion used in the verification (D23)-(D26). The correct first-order jump operators should carry √(γt) prefactors. This is not a mere notational slip: the recovery verification is the only analytical basis for the N=4 dephasing claim, so the derivation must be redone with consistent scaling.
  3. [Abstract; Appendix A, Eqs. (1)-(3)] The claimed qudit encoding in arbitrary dimensions is not supported by the group-theoretic construction as written. For d>2, π(h)=U_BS exp(i2π/(Nd) a1†a1) U_BS† satisfies π(h)^{2N}=exp(i4π/d a1†a1), which is not the identity operator on the physical Hilbert space. Hence π is not a representation of the stated group G=⟨g,h|h^{2N}=e⟩. For d=2 the relation holds (since 4π/d=2π), which explains why the qubit examples work, but the arbitrary-dimension claim requires either redefining the group as ⟨g,h|h^{Nd}=e⟩ (for which π(h)^{Nd}=I) or explicitly stating that π is only required to reproduce the logical action on the codespace rather than being a full group representation.
  4. [Abstract and Conclusions; Fig. 2] The statement that the codes exhibit 'no trade-off between protection against dephasing and photon loss' is too strong. Fig. 2(b) shows the loss performance has an optimum near N=4 and degrades for larger N, similar to the single-mode case; the dephasing performance improves up to N=8 in Fig. 2(a). Thus the improvement is a removal of the dephasing side of the trade-off in the studied range, not a removal of the trade-off itself. The claim should be qualified to the range where loss performance is comparable to the single-mode code.
minor comments (5)
  1. [Appendix C, subsection 'K=2, N=4 binomial encoding'] The statement 'It can be verified that the Knill-Laflamme conditions for the loss channel, up to first order, are satisfied' omits the verification entirely. Since the N=4 loss recovery is one of the paper's main analytical results, the calculation should be shown or at least summarized.
  2. [Appendix D, Eq. (D18)] The projection P_E is defined using an unspecified state |E⟩, and the unitary in Eq. (D19) depends on it. The recovery map is therefore not completely specified. Please define |E⟩ explicitly and show that the states reachable by the first-order dephasing errors are in the support of P_L or P_E as claimed.
  3. [Fig. 2 and numerical methods] No details are given for the SDP optimization (Hilbert-space truncation, solver, convergence tolerance), and no code is provided. Given that the dephasing results for N=6 and N=8 rest entirely on the numerics, please include these details or a link to a reproducible implementation.
  4. [Throughout the text] There are several typographical and formatting artifacts, e.g. the garbled axis labels in Fig. 2 ('0 - 2 - 4t w o - m o d e') and inconsistent uses of 'Knill-Lafflame' versus 'Knill-Laflamme'. A careful proofreading pass is needed.
  5. [Appendix E, Eq. (E9)] The discussion around Eq. (E9) correctly notes that the diagonal KL entries for the continuous dephasing channel are unequal for finite N, with equality only for N→∞ or correlated dephasing. To avoid confusion with the first-order KL claim, please state explicitly that this inequality concerns the full (non-truncated) channel and does not by itself contradict a first-order KL analysis.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity; the group-theoretic construction, analytic recovery maps, and SDP benchmarks are self-contained, and the same-group citations are contextual.

full rationale

Central derivation is self-contained: Appendix A obtains the codeword support (Eq. (3)) by imposing the representation conditions π(h)|k_N> = ω_d^k |k_N> and π(g)|k_N> = |(k⊕1)_N> on a general superposition; the binomial instances (Eqs. (6)-(7)) then fix the free coefficients f_mn and are not assumed to be optimal. The dephasing/loss benchmarks use SDP-optimal recovery (Ref. [27]) and analytic first-order recovery maps (Appendices C-D) with explicit KL checks against the same noise operators; the encoding angles δ=π/4,3π/4, φ=π/(2N) are chosen by analytically and numerically minimizing dephasing infidelity, which is parameter optimization, not a fitted input renamed as a prediction. The exact-correlated-dephasing claim is proved in Appendix G by explicit commutation of the CX gates with e^{iφ(n1+n2)} and by the action of the CX on the codeword Fock states; the proof does not presuppose the recovery. The paper's same-group citations (Refs. [18,28,42]) are contextual (single-mode RSB qudit definition, telecorrection distinguishability bound, teleportation circuit) and are not load-bearing. The manuscript itself flags a limitation: Appendix E, Eq. (E9), shows the KL diagonal entries for independent Gaussian dephasing are unequal for finite N, so exact full-channel recovery is not claimed; this is a correctness caveat, not circularity. Overall, no step in the claimed derivation reduces by construction to its own input.

Assumptions & free parameters 3 free parameters · 5 assumptions · 0 invented entities

The central construction is mathematically self-contained given standard Fock-space algebra and the group-theoretic framework of Ref. [14]. The free parameters are the beam-splitter angles, the superposition coefficients, and the instance parameters; the beam-splitter angles are optimized against the same noise used in the benchmarks, which is the main reason the no-trade-off claim should be read as a design result rather than an independent prediction. No new physical entities are introduced.

free parameters (3)
  • Beam-splitter angles (delta,phi) for two-mode code = delta=pi/4,3pi/4, phi=pi/(2N) for optimal dephasing
    These angles determine the U_BS rotation in Eqs. (4)-(7); the optimal values are selected by requiring dephasing KL conditions and minimizing dephasing infidelity, and the performance claims depend on this choice.
  • Superposition coefficients f_mn in codewords = f_00=f_10=1/sqrt(2), all others 0 for the K=2 binomial instance
    The general codeword family Eq. (3) leaves f free; the benchmarks choose the binomial coefficients, which defines the specific code analyzed.
  • Symmetry order N and mode number d = d=2, N=2,4,6,8 in benchmarks
    These are instance parameters of the family, not fitted to data, but the numerical no-trade-off claim is only demonstrated for these values.
assumptions (5)
  • domain assumption The group-theoretic encoding framework of Denys and Leverrier (Ref. [14]) is valid: a homomorphism lambda from group G to the logical Pauli group and a physical representation pi implementing lambda on a subspace yield a codespace with the desired logical gates.
    The paper builds directly on this framework without reproving it; if the framework fails, the codeword derivation in Appendix A collapses.
  • standard math Canonical commutation relations and Fock-space algebra, including beam-splitter transformation relations such as the one used to derive Eq. (A3).
    The codeword derivation in Appendix A relies on these standard identities without proving them.
  • domain assumption The modular number measurement POVM {P_l tensor P_m} is a physically implementable, non-destructive syndrome measurement for the codes.
    The recovery maps in Appendices C and D assume this measurement is available and projective; the paper cites Refs. [13,15] for it rather than proving implementability.
  • domain assumption The correlated dephasing noise model in Eq. (11), with Hamiltonian H(t)=nu c(t)(n1+n2) and stochastic c(t), describes the target environment.
    The exact-correction result in Appendix G is proven for this specific channel; it does not cover other multi-mode dephasing correlations.
  • domain assumption For benchmarking, loss and dephasing rates are set equal in both modes and the channels are treated as independent and Markovian.
    The performance comparisons and the claim of improved dephasing protection are generated under this symmetric local noise model.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Multimode rotationally symmetric bosonic codes from group-theoretic construction." pith.science (2026). https://pith.science/paper/LABI5DML

@misc{pith2026250820647,
  author       = {Pith},
  title        = {Pith review of: Multimode rotationally symmetric bosonic codes from group-theoretic construction},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/LABI5DML}},
  note         = {Machine review of arXiv:2508.20647}
}
read the original abstract

We introduce a new family of multi-mode, rotationally symmetric bosonic codes inspired by the group-theoretic framework of [Phys. Rev. Lett. 133, 240603 (2024)]. Such a construction inverts the traditional paradigm of code design by identifying codes from the requirement that a group of chosen logical gates should be implemented by means of physically simple logical operations, such as linear optics. Leveraging previously unexplored degrees of freedom within this framework, our construction preserves rotational symmetry across multiple modes, enabling linear-optics implementation of the full Pauli group. These codes exhibit improved protection against dephasing noise, outperforming both single-mode analogues and earlier multi-mode constructions. Notably, they allow exact correction of correlated dephasing and support qudit encoding in arbitrary dimensions. We analytically construct and numerically benchmark two-mode binomial codes instances, and demonstrate that, unlike single-mode rotationally symmetric bosonic codes, these exhibit no trade-off between protection against dephasing and photon loss.

Figures

Figures reproduced from arXiv: 2508.20647 by the authors.

Figure 1
Figure 1. FIG. 1: Application of the logical [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: Performance of the simple instances in Eqs.( [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: Exact error-correcting circuit for correlated [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
Figures from the paper (6 more)
Figure 4
Figure 4. Figure 4: FIG. 4: An encoding and error-correcting procedure for the rotation-symmetric codes on two modes. [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: (a) Entanglement infidelity as a function of [PITH_FULL_IMAGE:figures/full_fig_p016_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6 [PITH_FULL_IMAGE:figures/full_fig_p017_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7 [PITH_FULL_IMAGE:figures/full_fig_p018_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8: A comparison between the optimal performance of the single-mode binomial code, [PITH_FULL_IMAGE:figures/full_fig_p019_8.png]
Figure 9
Figure 9. Figure 9: FIG. 9: Implementation of the (a) Hadamard gate and (b) [PITH_FULL_IMAGE:figures/full_fig_p021_9.png]

Review history (2 revisions) →

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 36 canonical work pages

  1. [1]

    V . V . Albert, inQuantum Fluids of Light and Matter(IOS Press,

    work page

  2. [2]

    Z. Ni, S. Li, X. Deng, Y . Cai, L. Zhang, W. Wang, Z.-B. Yang, H. Yu, F. Yan, S. Liu,et al., Nature616, 56 (2023)

    work page 2023

  3. [3]

    Sivak, A

    V . Sivak, A. Eickbusch, B. Royer, S. Singh, I. Tsioutsios, S. Ganjam, A. Miano, B. Brock, A. Ding, L. Frunzio,et al., Nature616, 50 (2023)

    work page 2023

  4. [4]

    D. K. Tuckett, S. D. Bartlett, and S. T. Flammia, Phys. Rev. Lett.120, 050505 (2018)

    2018

  5. [5]

    Zhang, Y .-C

    J. Zhang, Y .-C. Wu, and G.-P. Guo, Physical Review A107 (2023), 10.1103/physreva.107.062408

    doi:10.1103/physreva.107.062408 2023

  6. [6]

    Noh and C

    K. Noh and C. Chamberland, Phys. Rev. A101, 012316 (2020)

    work page 2020

  7. [7]

    K. Noh, C. Chamberland, and F. G. Brandão, PRX Quantum3, 010315 (2022)

    work page 2022

  8. [8]

    Fukui, A

    K. Fukui, A. Tomita, A. Okamoto, and K. Fujii, Phys. Rev. X 8, 021054 (2018)

    work page 2018

Show all 47 references
  1. [9]

    Raveendran, N

    N. Raveendran, N. Rengaswamy, F. Rozp˛ edek, A. Raina, L. Jiang, and B. Vasi ´c, Quantum6, 767 (2022)

    2022

  2. [10]

    Berent, T

    L. Berent, T. Hillmann, J. Eisert, R. Wille, and J. Roffe, PRX Quantum5, 020349 (2024)

    2024

  3. [11]

    D. Ruiz, J. Guillaud, A. Leverrier, M. Mirrahimi, and C. Vuil- lot, Nature Communications16, 1040 (2025)

    2025

  4. [12]

    Gouzien, D

    É. Gouzien, D. Ruiz, F.-M. Le Régent, J. Guillaud, and N. San- gouard, Physical Review Letters131, 040602 (2023)

    2023

  5. [13]

    A. L. Grimsmo, J. Combes, and B. Q. Baragiola, Physical Re- view X10(2020), 10.1103/physrevx.10.011058

    2020 doi

  6. [14]

    Denys and A

    A. Denys and A. Leverrier, Phys. Rev. Lett.133, 240603 (2024)

    2024

  7. [15]

    M. H. Michael, M. Silveri, R. Brierley, V . V . Albert, J. Salmile- hto, L. Jiang, and S. M. Girvin, Physical Review X6, 031006 (2016)

    2016

  8. [16]

    Even though the physical representation is given using the pas- sive linear operations on thedmodes, note its difference from the one given in [14], which assumes a physical representation that is isomorphic to the Pauli group itself while the one pro- posed here is more general

  9. [17]

    Bergmann and P

    M. Bergmann and P. van Loock, Phys. Rev. A94, 012311 (2016)

    2016

  10. [18]

    García-Álvarez, C

    L. García-Álvarez, C. Calcluth, A. Ferraro, and G. Ferrini, Phys. Rev. Res.2, 043322 (2020)

    2020

  11. [19]

    Y . Xu, Y . Wang, and V . V . Albert, Phys. Rev. A110, 022402 (2024)

    2024

  12. [20]

    Descamps, A

    E. Descamps, A. Keller, and P. Milman, Phys. Rev. Lett.132, 170601 (2024)

    2024

  13. [21]

    Knill, R

    E. Knill, R. Laflamme, and G. J. Milburn, Nature409, 46 (2001)

    2001

  14. [22]

    Guillaud and M

    J. Guillaud and M. Mirrahimi, Physical Review X9, 041053 (2019)

    2019

  15. [23]

    V . V . Albert, S. O. Mundhada, A. Grimm, S. Touzard, M. H. Devoret, and L. Jiang, Quantum Science and Technology4, 035007 (2019)

    2019

  16. [24]

    Bosonic quantum error correc- tion with microwave cavities for quantum repeaters,

    S. S. Chelluri, S. Sharma, F. Schmidt, S. V . Kusmin- skiy, and P. van Loock, “Bosonic quantum error correc- tion with microwave cavities for quantum repeaters,” (2025), arXiv:2503.21569 [quant-ph]

    2025

  17. [25]

    Furthermore, our codes are also distinct from the bosonic quan- tum Fourier codes based on cat states along a similar group- theoretic construction [41]

  18. [26]

    I. L. Chuang, D. W. Leung, and Y . Yamamoto, Phys. Rev. A 6 56, 1114 (1997)

    1997

  19. [27]

    A. S. Fletcher, P. W. Shor, and M. Z. Win, Physical Review A 75(2007), 10.1103/physreva.75.012338

    2007 doi

  20. [28]

    Udupa, T

    A. Udupa, T. Hillmann, R. G. Ahmed, A. Smirne, and G. Fer- rini, arXiv preprint arXiv:2505.08670 (2025)

    2025 arXiv

  21. [29]

    Leonhardt, J

    U. Leonhardt, J. A. Vaccaro, B. Böhmer, and H. Paul, Phys. Rev. A51, 84 (1995)

    1995

  22. [30]

    Bergli, Y

    J. Bergli, Y . M. Galperin, and B. L. Altshuler, Phys. Rev. B74, 024509 (2006)

    2006

  23. [31]

    Bergli, Y

    J. Bergli, Y . M. Galperin, and B. L. Altshuler, New Journal of Physics11, 025002 (2009)

    2009

  24. [32]

    Schriefl, Y

    J. Schriefl, Y . Makhlin, A. Shnirman, and G. Schön, New Jour- nal of Physics8, 1 (2006)

    2006

  25. [33]

    Müller, J

    C. Müller, J. H. Cole, and J. Lisenfeld, Reports on Progress in Physics82, 124501 (2019)

    2019

  26. [34]

    Benedetti, M

    C. Benedetti, M. G. Paris, and S. Maniscalco, Physical Review A89, 012114 (2014)

    2014

  27. [35]

    Lescanne, M

    R. Lescanne, M. Villiers, T. Peronnin, A. Sarlette, M. Delbecq, B. Huard, T. Kontos, M. Mirrahimi, and Z. Leghtas, Nature Physics16, 509 (2020)

    2020

  28. [36]

    B. J. Chapman, S. J. de Graaf, S. H. Xue, Y . Zhang, J. Teoh, J. C. Curtis, T. Tsunoda, A. Eickbusch, A. P. Read, A. Koot- tandavida, S. O. Mundhada, L. Frunzio, M. Devoret, S. Girvin, and R. Schoelkopf, PRX Quantum4, 020355 (2023)

    2023

  29. [37]

    Royer, S

    B. Royer, S. Singh, and S. Girvin, PRX Quantum3, 010335 (2022)

    2022

  30. [38]

    Suppressing logical errors with multimode quantum error correction,

    Nordquantique, “Suppressing logical errors with multimode quantum error correction,” (2025)

    2025

  31. [39]

    Hillmann, F

    T. Hillmann, F. Quijandría, G. Johansson, A. Ferraro, S. Gas- parinetti, and G. Ferrini, Physical review letters125, 160501 (2020)

    2020

  32. [40]

    A. M. Eriksson, T. Sépulcre, M. Kervinen, T. Hillmann, M. Ku- dra, S. Dupouy, Y . Lu, M. Khanahmadi, J. Yang, C. Castillo- Moreno,et al., Nature Communications15, 2512 (2024)

    2024

  33. [41]

    Bosonic quantum fourier codes,

    A. Leverrier, “Bosonic quantum fourier codes,” (2025), arXiv:2505.16618 [quant-ph]

    2025

  34. [42]

    Hillmann, F

    T. Hillmann, F. Quijandría, A. L. Grimsmo, and G. Ferrini, PRX Quantum3, 020334 (2022). Appendix A: Multimode rotation-symmetric codewords In this section, we provide a derivation of the qudit codewords Eq. (3). Firstly, note that unlike [14] where both the groupG and the phys...

    2022

  35. [43]

    ˆPC + ˆU ˆPE( ˆU00( ˆP0⊗ ˆP0) ˜N(ˆρ)(ˆP0⊗ ˆP0) ˆU†

  36. [44]

    We have shown that an exact recovery map forK= 2,N= 4two-mode code up to the first order can be found if the encoding mode is fixed to beδ=π/4,3π/4(independenly onϕ)

    ˆPE ˆU† = ˆρ−10(γ1t+γ 2t)ˆρ+O((γit)2),(D23) where we have used ˆU00( ˆP0⊗ ˆP0) ˜N(ˆρ)(ˆP0⊗ ˆP0) ˆU† 00 = ˆρ−30(γ1t+γ 2t)ˆρ+1 4(γ1t+γ 2t)(ˆa† 1ˆa1 + ˆa† 2ˆa2)ˆρ(ˆa† 1ˆa1 + ˆa† 2ˆa2) +O((γit)2).(D24) The single-photon exchange correction gives ˆU13( ˆP1⊗ ˆP3) ˜N(ˆρ)(ˆP1⊗ ˆP3) ˆU...

  37. [45]

    Knill-Laflamme matrix for dual codewords under dephasing noise The effective noisy channel is given by ˜N= ˆU† BS N ˆUBS where the noise channelNcan be expanded in terms of the elements of the continuous Kraus operators, N∼{exp i ˆA(θ1,θ 2) =e iθ1ˆa† 1ˆa1+iθ2ˆa† 2ˆa2|θ 1,θ 2∈[...

  38. [46]

    Moreover, we also see that for allθ 1 =θ 2 corresponding to the case of correlated dephasing, the entries given in Eq.(E9) are exactly equal

    π 2N form∈Z, and taking the limitN→∞. Moreover, we also see that for allθ 1 =θ 2 corresponding to the case of correlated dephasing, the entries given in Eq.(E9) are exactly equal. This justifies the existence of the recovery circuit in Fig. 3. Note that the above discussion ho...

  39. [47]

    +” andα|+ M⟩−β|− M⟩for the outcome “−

    Distinguishing the dual codewords and understanding the phase trajectories In this Section, we will see how the two-mode code is distinct in performance under dephasing as compared to a single mode code in terms of the phase distinguishability of the codewords. Similar to the ...

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.