Pith. sign in

REVIEW 3 major objections 5 minor 60 references

Parity Mapping for Quantum Optimization on Frustrated Ising Rings

T0 review · 3 major / 5 minor · reviewed 2026-08-16 · deepseek-v4-flash

Pith's one-line read The parity mapping turns the frustrated Ising ring into an exactly solvable constant-depth QAOA instance and widens the annealing gap.

desk verdict Solid, genuinely novel analysis of parity QAOA on a sparse graph; the constant-depth exact-ground-state claim is a real result but is over-sold as exact and the QA gap comparison lacks a defined normalization. read the letter →

arxiv 2608.12213 v1 pith:6LANLNYF submitted 2026-08-12 quant-ph

classification quant-ph
keywords paritymappingfrustratedIsingringquantumannealingQAOAspectralgapground-statepreparationLHZarchitectureoptimization
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 asks whether the parity encoding of an Ising model—mapping each interaction $\hat\sigma^z_j\hat\sigma^z_{j+1}$ to a local field $\tilde J_k \tilde\sigma^z_k$ on a parity qubit—improves ground-state preparation in both quantum annealing (QA) and QAOA on the frustrated Ising ring, a sparse benchmark whose spectral gap closes exponentially with system size. It reports a spectral benefit in the continuous-time setting: under the energy normalization used in the paper, an ideal parity implementation with a single global constraint shows no exponential gap closing over the accessible sizes, while a hardware-motivated decomposition into local constraints restores exponential closing with roughly half the decay exponent ($e^{-0.51N}$ versus $e^{-0.96N}$). In the digitized setting, it reports a stronger structural result: the exact ground state of the ring and its symmetry-broken variants is obtained with unit probability at circuit depth equal to the number of distinct coupling values appearing in the Hamiltonian, with all constraint angles zero, so only single-qubit rotations are needed and depth does not grow with $N$. For a variant with equal coupling magnitudes and one antiferromagnetic bond, where the constraint unitary is necessary, Parity-QAOA reaches near-unit ground-state probability at constant depth and reduces the total CNOT count from $O(N^2)$ to $O(N)$ relative to standard QAOA. A sympathetic reader would take away that parity encoding can remove both the small-gap bottleneck and the layer growth of QAOA for this model.

What carries the argument

The load-bearing object is the parity mapping, which replaces every logical Ising coupling $J_j \hat\sigma^z_j\hat\sigma^z_{j+1}$ by a local field $\tilde J_k \tilde\sigma^z_k$ on a parity qubit, turning the ring Hamiltonian into $K=N$ independent fields plus a global parity constraint. In Parity-QAOA the unitary $\tilde U_z$ is a product of single-qubit rotations whose rates are set by the distinct $\tilde J_k$, and when the constraint unitary is omitted the ground-state probability factorizes as $\prod_k P_\uparrow(\theta^x,\tilde J_k\theta^z)$. The mechanism that carries the constant-depth result is frequency selectivity: qubits with different $|\tilde J_k|$ accumulate different phases under $\tilde U_z$, and the paper proves, by inverting the Vandermonde matrix arising from nested commutators of $\tilde H_x$ and $\tilde H_z$, that this gives arbitrary individual control over each qubit. The same argument shows the obstruction when two local fields have opposite signs and equal magnitude: the Vandermonde matrix loses invertibility, and the parity constraint unitary becomes necessary.

What would settle it

Compute the minimum gap $\Delta_{\min}$ for standard QA and Parity-QA with an explicitly fixed normalization, for example rescaling each interpolating Hamiltonian to the same operator norm or fixing the constraint strength $C$ relative to the total energy scale in a stated way, and check whether the parity gap remains larger and non-closing; if the amplification disappears under a natural convention, the claimed spectral benefit is an artifact of the undefined normalization.

Watch

Extended reading notes

Core claim

The central discovery is that the parity mapping changes the resource scaling of both continuous-time and digitized quantum optimization on the frustrated Ising ring. For Parity-QA, finite-size spectra show the minimum gap $\Delta_{\min}$ is larger than in standard linear-schedule QA; with no ancillas, the gap shows no exponential closing over $N=5,\ldots,15$, whereas adding the hardware-required $(N-3)/2$ ancillas to decompose the $K$-body constraint restores exponential closing with a fitted exponent $\Delta_{\min}/J\sim e^{-0.51N}$, compared with $e^{-0.96N}$ for standard QA. For Parity-QAOA, the exact ground state ($P_{\rm GS}=1$) is obtained at depth $P$ equal to the number of distinct $J_j$ values, with optimized constraint angles $\theta^c_p=0$; the protocol therefore uses only single-qubit $z$- and $x$-rotations and its depth is independent of system size. The paper proves the underlying mechanism: with distinct absolute coupling magnitudes $|\tilde J_k|$, the generated Lie algebra contains each individual qubit rotation, so qubits can be controlled separately; when $|\tilde J_k|=|\tilde J_l|$ with opposite signs, single-qubit rotations alone cannot separate the two qubits, and the constraint unitary becomes essential. In that equal-magnitude variant, Parity-QAOA still reaches $P_{\rm GS}>0.99$ at constant $P^*=4$ for all sizes studied, versus $P^*=\lceil N/2\rceil-1$ for standard QAOA, reducing total CNOT count from $O(N^2)$ to $O(N)$ at similar $O(N)$ total depth.

Load-bearing premise

The spectral-gap advantage of Parity-QA rests on an energy normalization that the abstract invokes but the main text never specifies, so a different way of setting the constraint strength or comparing the two Hilbert-space sizes could shrink or erase the reported gap amplification.

Editorial extensions

If this is right

  • For the original frustrated ring and its symmetry-broken variants, exact ground-state preparation with Parity-QAOA requires $P$ equal to the number of distinct coupling values, independent of the number of qubits $N$.
  • This improves on standard QAOA, which for exact preparation on the same model needs a number of layers growing quadratically with $N$.
  • In the equal-magnitude ring with one antiferromagnetic bond, Parity-QAOA reaches $P_{\rm GS}>0.99$ at constant depth $P^*=4$ for all sizes simulated, while standard QAOA needs $P^*=\lceil N/2\rceil-1$ layers; the total CNOT count drops from $O(N^2)$ to $O(N)$, with both protocols running at $O(N)$ total depth.
  • For continuous-time QA, ancilla-free Parity-QA shows no exponential gap closing over the accessible sizes, while the realistic ancilla-decomposed implementation closes as $e^{-0.51N}$, about half the standard exponent $e^{-0.96N}$, so the parity encoding preserves a quantifiable spectral benefit under local constraints.
  • The presence of the constraint unitary in Parity-QAOA is controlled by the coupling spectrum: it can be omitted when all $|\tilde J_k|$ are distinct and becomes necessary when opposite-sign equal-magnitude local fields appear.

Reading between the lines

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

  • A natural extension the paper does not pursue: the factorization argument suggests the same constant-depth phenomenon should occur in any sparse problem whose parity image has distinct local-field magnitudes, so the protocol is a candidate testbed for transferable QAOA parameters across sizes.
  • The gap amplification claim is normalization-sensitive; a reader should not infer from it that parity encoding generically removes small gaps in annealing until an unambiguous energy convention is stated.
  • The opposite-sign obstruction could be used deliberately: constructing instances with pairs of equal-magnitude opposite local fields produces a class where the constraint unitary is genuinely unavoidable, offering a controlled setting to benchmark constraint-enforcement hardware.
  • Because the constraint-free protocol uses only single-qubit rotations, an experimental implementation on any qubit platform could test the constant-depth claim at larger $N$ without requiring the multi-qubit constraint gates; the practical depth would be set by single-qubit gate count.
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

3 major / 5 minor

Summary. This paper studies the effect of the parity (LHZ) mapping on quantum annealing (QA) and QAOA for the frustrated Ising ring and two families of variants. For QA, finite-size numerical spectra are used to claim that the parity mapping enlarges the minimum spectral gap under an energy normalization that is referenced in the abstract but never defined in the main text; an ideal ancilla-free global constraint shows no exponential gap closing, while a hardware-motivated local-constraint decomposition restores exponential closing with a smaller fitted exponent. For QAOA, the central result is that on the original frustrated ring and its symmetry-broken variants, the exact ground state is obtained with unit probability at circuit depth equal to the number of distinct coupling values, using only single-qubit rotations and with the constraint unitary angles optimized to zero. This claim is supported by a factorized single-qubit expression for the ground-state probability (Appendix B) and a dynamical-Lie-algebra controllability argument based on Vandermonde matrices (Appendix C). For a modified ring with all coupling magnitudes equal to one and a single antiferromagnetic bond, the constraint unitary is required, and the paper reports that Parity-QAOA reaches ground-state probability above 0.99 with a constant number of layers, reducing total CNOT count from O(N^2) for standard QAOA to O(N).

Significance. If the main claims hold, the paper reports a striking advantage of the parity encoding on a sparse, frustrated model: zero qubit overhead, a constraint-free exact QAOA protocol whose required depth is independent of system size, and a CNOT-count reduction from O(N^2) to O(N) in the constraint-required variant. The factorized ground-state probability formula in Appendix B is a genuinely clean analytic result, and the Vandermonde-based controllability argument is elegant and machine-verifiable. The paper also provides an honest quantified comparison of the QA spectral-gap trade-off between ideal global constraints and hardware-realizable local constraints, and it makes its simulation code publicly available. However, two load-bearing points need attention: the energy normalization behind the QA gap-amplification claim is never defined, and the exactness of the constant-depth QAOA result rests on BFGS-optimized numerical angles rather than on a proof that the finite-depth parameters make each single-qubit factor exactly one.

major comments (3)
  1. [Sec. V A and Eq. (6)] The energy normalization used for the QA gap comparison is never defined. The abstract states that parity mapping increases the minimum spectral gap "under the energy normalization used in this work," but the main text compares the standard Hamiltonian ॐH_z of Eq. (2) with the parity Hamiltonian ॐH_targ = ॐH_z + C ॐH_c of Eq. (6) using C = J (or C = 2J) without specifying how the different Hilbert-space dimensions (N vs. K = N + n_a) and the added constraint-energy scale are normalized. Without such a normalization, the reported gap amplification in Fig. 3 is not a well-defined property of the parity mapping. Please define the normalization explicitly and either re-plot the gaps under that normalization or qualify the claim accordingly.
  2. [Sec. V B 1 and Appendices B/C] The system-size-independent exact-preparation claim is not rigorously supported. The factorization in Eq. (B4)–(B5) is exact, but the claim P_GS = 1 for all N requires each single-qubit factor P↑(θx, ॐJ_k θz) to be exactly one at the chosen depth P = M. The parameters in Table I are obtained by BFGS optimization, which cannot certify exact equality; if each factor is actually 1 − ε with ε > 0, then P_GS = (1−ε)^{N−3} ≈ e^{−ε(N−3)}, which decays exponentially with N and invalidates the claimed size independence. Appendix C proves controllability via the dynamical Lie algebra with arbitrarily long pulse sequences, and the text itself acknowledges that it does not prove that P = M layers suffice. Please provide either an analytic construction or a certified numerical proof that the Table I parameters give P↑ = 1 exactly for each distinct ॐJ_k, or state the result as numerically exact only and revise the "unit probability" and "P_GS = 1 for N > 10^6" claims accordingly.
  3. [Sec. V B 3] The comparison between standard QAOA and Parity-QAOA for the uniform-coupling variant is presented as a resource advantage based on P_GS > 0.99, but the two protocols are not compared at the same target accuracy. Standard QAOA is reported to reach the exact ground state at P* = ⌈N/2⌉−1, whereas Parity-QAOA only reaches P_GS > 0.99 at P = 4 and is not shown to reach P_GS = 1 at any simulated depth. Since the residual error is still decreasing with depth for Parity-QAOA, the claimed constant-depth resource advantage is tied to the chosen 0.99 threshold. Please clarify whether the O(N) vs. O(N^2) CNOT-count comparison is meant for exact preparation or for fixed finite accuracy, and if the latter, state the accuracy threshold explicitly in the comparison.
minor comments (5)
  1. [Sec. II and Eq. (3)] The notation K is introduced as the number of parity qubits with K = N for the ring, but later K = N + n_a is used when ancillas are added. Please make the distinction between logical N and physical K consistent throughout the text.
  2. [Appendix C, Eqs. (C1)–(C3)] The Vandermonde argument jumps from a sum over K physical qubits to a matrix acting on M distinct ॐJ_k without explicitly stating that qubits with equal local fields are grouped. Adding a sentence on grouping qubits with identical ॐJ_k would make the derivation easier to follow.
  3. [Table I] The table reports angles with many decimal places but no optimization tolerance or residual value. Please state the BFGS stopping criterion and the resulting 1 − P_GS at N = 7 for the reported parameters.
  4. [Sec. V B 2] The sentence "transferring parameters from smaller to larger N seems necessary to navigate barren plateaus" is speculative; the paper does not perform a barren-plateau analysis. Please either cite a specific analysis or soften the statement.
  5. [Fig. 3] The log-linear fits for the exponential gap closing are based on only five data points (N = 5, 7, ..., 15) and are quoted as ≈e^{−0.96N} and ≈e^{−0.51N} without confidence intervals. Please include fit uncertainties or explicitly state the small sample size.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central constant-depth Parity-QAOA result follows from an explicit single-qubit factorization, and no prediction reduces to its own input.

full rationale

The paper's headline results are not circular. The exact-ground-state Parity-QAOA claim is built on Eq. (B5), P_GS = P_up^{N-3}(θx,Jθz) P_up^2(θx,Jwθz) P_up(θx,−Jfθz), a closed-form single-qubit factorization valid for arbitrary N. The N-dependence enters only through powers of the single-qubit probabilities, so if a parameter set makes each factor equal to 1, the same parameters succeed for every N. The parameters in Table I are obtained by numerical optimization, not by using the target conclusion as an input; the claim that they achieve exactly P_GS = 1 is a numerical robustness concern rather than a circularity. The parity-QA gap comparison uses an energy normalization that is not explicitly defined in the main text, which is a limitation, but the gap numbers are computed directly from the interpolating Hamiltonians and are not fitted to the conclusion. Self-citations to the LHZ/parity architecture and to the same group's earlier QAOA studies are contextual baselines; no load-bearing derivational step is justified solely by a self-citation. Appendix C's controllability argument is an independent Vandermonde-matrix proof. Thus no claimed prediction is equivalent to its inputs by construction.

Assumptions & free parameters 2 free parameters · 8 assumptions · 0 invented entities

No new physical entities are introduced. The parity mapping and ancilla qubits are existing architecture elements from prior work. The central claims rest on standard adiabatic theorem assumptions, a prior result about the frustrated ring's gap, the LHZ constraint counting rule, and algebraic controllability arguments.

free parameters (2)
  • Constraint strength C for original ring = C=J
    Chosen to satisfy the analytic condition C > min |J_j| = 0.45J from Ref. [39]. It affects the spectral gap magnitudes and the QAOA cost function, but it is not fitted to the reported scalings.
  • Constraint strength C for +/-1 ring = C=2J
    Chosen because min |J_j| = J in the uniform-magnitude variant; again follows the analytic condition rather than data fitting.
assumptions (8)
  • domain assumption The minimum spectral gap along a linear schedule determines the adiabatic time cost of QA.
    Standard QA performance criterion used throughout Section V A; relies on the adiabatic theorem.
  • domain assumption The frustrated Ising ring with 0 < J_f < J_w < J and JJ_f > J_w^2 has an exponentially closing gap.
    Result from Ref. [37]; the paper uses it as the benchmark's difficulty premise, with parameters from Refs. [14,25].
  • standard math The parity mapping requires K-N+1 constraints; for the ring one global constraint is sufficient.
    From the LHZ construction in Ref. [27] and the paper's Eqs. (4)-(5).
  • domain assumption A constraint strength C > min_j |J_j| guarantees the ground state of the parity target Hamiltonian is a valid logical state.
    The paper cites Ref. [39] for this condition and sets C=J or C=2J accordingly.
  • standard math Individual single-qubit controllability follows from invertibility of the Vandermonde matrices when the |J_k| are distinct.
    Appendix C proves this using the dynamical Lie algebra and Vandermonde determinant, citing Ref. [53].
  • standard math For qubits with opposite local fields J and -J, no sequence of common single-qubit rotations can align both to the same target state.
    Appendix C proof using the identity sigma_x sigma_z sigma_x = -sigma_z; used to explain why the constraint unitary is required for the +/-1 ring.
  • domain assumption The ground state of the original frustrated ring maps to the all-parity-qubits-up state in the parity encoding.
    Used in Appendix B and Section V B 1; follows from the bit-to-spin convention where 0 maps to |up>.
  • domain assumption The variational parameter optimizations with 100 random restarts and BFGS find the relevant optima.
    Appendix A; QAOA performance depends on best-found parameters, and the paper provides no global optimality certificate.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Parity Mapping for Quantum Optimization on Frustrated Ising Rings." pith.science (2026). https://pith.science/paper/6LANLNYF

@misc{pith2026260812213,
  author       = {Pith},
  title        = {Pith review of: Parity Mapping for Quantum Optimization on Frustrated Ising Rings},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6LANLNYF}},
  note         = {Machine review of arXiv:2608.12213}
}
read the original abstract

The frustrated Ising ring is one of the simplest models exhibiting exponential closing spectral gaps, making it a paradigmatic and challenging benchmark for quantum annealing (QA). Ground-state preparation for this model has therefore been studied extensively in both continuous-time QA and digitized protocols such as the Quantum Approximate Optimization Algorithm (QAOA). Here, we use the frustrated Ising ring to investigate how the parity mapping affects the performance of both QA and QAOA. For QA, finite-size calculations show that the parity mapping increases the minimum spectral gap under the energy normalization used in this work, thereby enabling faster continuous-time ground state preparation protocols. An ideal implementation of Parity-QA, with a single global constraint, shows no evidence of exponential gap closing over the accessible system sizes, whereas a hardware-motivated decomposition into local constraints restores the exponential decrease, albeit with a smaller fitted exponent than conventional QA. For the digitized protocol, we find that the number of Parity-QAOA layers required to prepare the exact ground state remains constant over the simulated sizes, improving upon the quadratic scaling required by conventional QAOA. To investigate the role of constraints in Parity-QAOA, we further consider a modified Ising ring instance in which the constraint term is essential for preparing the target ground state. We then compare the corresponding resource requirements with those of conventional QAOA.

Figures

Figures reproduced from arXiv: 2608.12213 by the authors.

Figure 1
Figure 1. FIG. 1: The Ising ring with N = 7 in its logical representation (a) and parity representations (c)–(d). The parity transformation [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2: The frustrated Ising ring with N spins. The antiferromagnetic bond [PITH_FULL_IMAGE:figures/full_fig_p005_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3: Minimal energy gap ∆ [PITH_FULL_IMAGE:figures/full_fig_p007_3.png] view at source ↗
Figures from the paper (5 more)
Figure 4
Figure 4. Figure 4: FIG. 4: Ground state infidelity 1 [PITH_FULL_IMAGE:figures/full_fig_p008_4.png]
Figure 5
Figure 5. Figure 5: FIG. 5: The probability of being in the exact ground state [PITH_FULL_IMAGE:figures/full_fig_p009_5.png]
Figure 6
Figure 6. Figure 6: FIG. 6: Performance and resource scaling of standard and Parity-QAOA on the modified frustrated ring ( [PITH_FULL_IMAGE:figures/full_fig_p010_6.png]
Figure 7
Figure 7. Figure 7: FIG. 7: Best found Parity-QAOA parameters ( [PITH_FULL_IMAGE:figures/full_fig_p010_7.png]
Figure 8
Figure 8. Figure 8: FIG. 8: Best found QAOA parameters [PITH_FULL_IMAGE:figures/full_fig_p014_8.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

60 extracted references · 40 canonical work pages

  1. [1]

    " 0 ="# 1 =#

    An example of a ring withN= 7 is shown in Fig. 1 (a). (a) 0 1 2 345 6 01 12 23 344556 06 ="" 0 ="# 1 =#" 1 =## 0 ij ij 01 12 23 344556 06 05 14 Rz Rz Rz Rz Rz Rz Rz Rz Rz Rz Rz (e) (b) (c) (f) (d) (g) FIG. 1: The Ising ring with N = 7 in its logical representation (a) and parity representations (c)–(d). The parity transformation table is given in (b). Wit...

  2. [2]

    The alternating unitary operators are defined using the driving Hamiltonian and the target Hamiltonian, respectively

    ˆUz(θz 1)|ψ0⟩(7) where the initial state|ψ 0⟩=|+⟩ ⊗N = |↑⟩+|↓⟩√ 2 ⊗N . The alternating unitary operators are defined using the driving Hamiltonian and the target Hamiltonian, respectively. At each layerp= 1,···,P, we have: ˆUx(θx p) = e−iθx p ˆHx , ˆUz(θz p) = e−iθz p ˆHtarg = e−iθz p ˆHz .(8) The 2P parameters (θx,θz) are optimized by minimizing the vari...

  3. [3]

    frustrated

    ˜Uz(θz 1)| ˜ψ0⟩.(10) The initial state is| ˜ψ0⟩=|+⟩ ⊗K. Three types of unitaries are applied to the quantum state at each layerp, and this is due to the additional constraint term ˜Hc in the target Hamiltonian ˜Htarg. These unitaries read: ˜Ux(θx p) = e−iθx p ˜Hx , ˜Uc(θc p) = e−iθc p ˜Hc , ˜Uz(θz p) = e−iθz p ˜Hz .(11) At each layer, there are three free...

  4. [4]

    Improved Controllability Parity-QAOA simulations were performed on the original frustrated ring and on the modified cases obtained by (i) settingJ 1 =J 2 = 0.8 and (ii) further settingJ N−1 = 0.85; see Sec. IV. There are three distinctJ j values in the original case, four in case (i), and five in case (ii). For N = 7, the ground-state probabilities are su...

  5. [5]

    Optimal control parameters and transferability We observe for small system sizes, N = 7, that Parity-QAOA consistently selects trivial constraint unitaries, with θc p = 0, for the frustrated Ising model. We can classically simulate the single-qubit rotations ˜Ux(θx p) and ˜Uz(θz p) to obtain a set of optimized parameters (θ x,θz) that drive the system to ...

  6. [6]

    As described in Sec

    Use of Constraints We now turn to the regime in which the constraint unitary is indispensable. As described in Sec. IV, we set all coupling magnitudes to unity:J j/J= 1 for all couplings except for a single coupling,J f/J=−1. In the parity representation, the ground state is characterized by all qubits aligning with their local fields ˜Jk, except for a si...

  7. [7]

    Nielsen and I

    M. Nielsen and I. L. Chuang,Quantum Computation and Quantum Information(Cambridge University Press, 2000). 15

  8. [8]

    Montanaro, npj Quantum Information2, 1 (2016)

    A. Montanaro, npj Quantum Information2, 1 (2016)

Show all 60 references
  1. [9]

    Alexeev, D

    Y. Alexeev, D. Bacon, K. R. Brown, R. Calderbank, L. D. Carr, F. T. Chong, B. DeMarco, D. Englund, E. Farhi, B. Fefferman, A. V. Gorshkov, A. Houck, J. Kim, S. Kimmel, M. Lange, S. Lloyd, M. D. Lukin, D. Maslov, P. Maunz, C. Monroe, J. Preskill, M. Roetteler, M. J. Savage, and...

  2. [10]

    Preskill, Quantum2, 79 (2018)

    J. Preskill, Quantum2, 79 (2018)

  3. [11]

    Bharti, A

    K. Bharti, A. Cervera-Lierta, T. H. Kyaw, T. Haug, S. Alperin-Lea, A. Anand, M. Degroote, H. Heimonen, J. S. Kottmann, T. Menke, W.-K. Mok, S. Sim, L.-C. Kwek, and A. Aspuru-Guzik, Rev. Mod. Phys.94, 015004 (2022)

  4. [12]

    A. B. Finnila, M. A. Gomez, C. Sebenik, C. Stenson, and J. D. Doll, Chemical Physics Letters219, 343 (1994)

  5. [13]

    Kadowaki and H

    T. Kadowaki and H. Nishimori, Physical Review E58, 5355 (1998)

  6. [14]

    G. E. Santoro, R. Martonak, E. Tosatti, and R. Car, Science295, 2427 (2002)

  7. [15]

    G. E. Santoro and E. Tosatti, Journal of Physics A: Mathematical and General39, R393 (2006)

  8. [16]

    Albash and D

    T. Albash and D. A. Lidar, Rev. Mod. Phys.90, 015002 (2018)

  9. [17]

    Jansen, M.-B

    S. Jansen, M.-B. Ruskai, and R. Seiler, Journal of Mathematical Physics48(2007)

  10. [18]

    Messiah,Quantum mechanics(Courier Corporation, 2014)

    A. Messiah,Quantum mechanics(Courier Corporation, 2014)

  11. [19]

    Roland and N

    J. Roland and N. J. Cerf, Phys. Rev. A65, 042308 (2002)

  12. [20]

    Cˆ ot´ e, F

    J. Cˆ ot´ e, F. Sauvage, M. Larocca, M. Jonsson, L. Cincio, and T. Albash, Quantum Science and Technology8, 045033 (2023)

  13. [21]

    Grabarits, F

    A. Grabarits, F. Balducci, and A. del Campo, PRX Quantum7, 010322 (2026), arXiv:2410.02520

  14. [22]

    Farhi, J

    E. Farhi, J. Goldstone, and S. Gutmann, A Quantum Approximate Optimization Algorithm (2014), arXiv:1411.4028 [quant-ph]

  15. [23]

    Zhou, S.-T

    L. Zhou, S.-T. Wang, S. Choi, H. Pichler, and M. D. Lukin, Physical Review X10, 021067 (2020)

  16. [24]

    Blekos, D

    K. Blekos, D. Brand, A. Ceschini, C.-H. Chou, R.-H. Li, K. Pandya, and A. Summer, Physics Reports1068, 1 (2024), a review on Quantum Approximate Optimization Algorithm and its variants

  17. [25]

    Cerezo, A

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio, and P. J. Coles, Nature Reviews Physics3, 625 (2021)

  18. [26]

    D’Alessandro,Introduction to quantum control and dynamics(Chapman and Hall/CRC, 2007)

    D. D’Alessandro,Introduction to quantum control and dynamics(Chapman and Hall/CRC, 2007)

  19. [27]

    Z.-C. Yang, A. Rahmani, A. Shabani, H. Neven, and C. Chamon, Phys. Rev. X7, 021027 (2017)

  20. [28]

    G. B. Mbeng, R. Fazio, and G. E. Santoro, Optimal quantum control with digitized quantum annealing (2019), arXiv:1911.12259 [quant-ph]

  21. [29]

    L. T. Brady, C. L. Baldwin, A. Bapat, Y. Kharkov, and A. V. Gorshkov, Phys. Rev. Lett.126, 070505 (2021)

  22. [30]

    V. R. Arezzo, R. Wang, K. Thengil, G. Pecci, and G. Santoro, Physical Review A113, 012610 (2026), arXiv:2509.17754

  23. [31]

    R. Wang, V. R. Arezzo, K. Thengil, G. Pecci, and G. E. Santoro, Quantum Science and Technology10, 035052 (2025)

  24. [32]

    Lloyd and S

    S. Lloyd and S. Montangero, Physical review letters113, 010502 (2014)

  25. [33]

    Lechner, P

    W. Lechner, P. Hauke, and P. Zoller, Science advances1, e1500838 (2015)

  26. [34]

    Ender, R

    K. Ender, R. ter Hoeven, B. E. Niehoff, M. Drieb-Sch¨ on, and W. Lechner, Quantum7, 950 (2023)

  27. [35]

    Lechner, IEEE Transactions on Quantum Engineering1, 1 (2020)

    W. Lechner, IEEE Transactions on Quantum Engineering1, 1 (2020)

  28. [36]

    Susa and H

    Y. Susa and H. Nishimori, Phys. Rev. A103, 022619 (2021)

  29. [37]

    Lanthaler, C

    M. Lanthaler, C. Dlaska, K. Ender, and W. Lechner, Phys. Rev. Lett.130, 220601 (2023)

  30. [38]

    Weidinger, G

    A. Weidinger, G. B. Mbeng, and W. Lechner, Physical Review A108, 032408 (2023)

  31. [39]

    Weidinger, G

    A. Weidinger, G. B. Mbeng, M. Fellner, D. Khachatryan, and W. Lechner, New Journal of Physics28, 034514 (2026), arXiv:2409.14786

  32. [40]

    Fellner, K

    M. Fellner, K. Ender, R. ter Hoeven, and W. Lechner, Quantum7, 952 (2023)

  33. [41]

    Wybo and M

    E. Wybo and M. Leib, Quantum8, 1554 (2024)

  34. [42]

    Bennett, N

    J. Bennett, N. Chancellor, V. Kendon, and W. Lechner, Phys. Rev. A112, 042616 (2025)

  35. [43]

    Roberts, L

    D. Roberts, L. Cincio, A. Saxena, A. Petukhov, and S. Knysh, Phys. Rev. A101, 042317 (2020)

  36. [44]

    G. B. Mbeng, A. Russomanno, and G. E. Santoro, SciPost Phys. Lect. Notes , 82 (2024)

  37. [45]

    Lanthaler and W

    M. Lanthaler and W. Lechner, New Journal of Physics23, 083039 (2021)

  38. [46]

    Mari´ c, S

    V. Mari´ c, S. M. Giampaolo, D. Kui´ c, and F. Franchini, New Journal of Physics22, 083024 (2020)

  39. [47]

    Mari´ c, S

    V. Mari´ c, S. M. Giampaolo, and F. Franchini, Communications Physics3, 220 (2020)

  40. [48]

    Torre, V

    G. Torre, V. Mari´ c, F. Franchini, and S. M. Giampaolo, Phys. Rev. B103, 014429 (2021)

  41. [49]

    Miyazaki and T

    R. Miyazaki and T. Yamamoto, Phys. Rev. A111, 062612 (2025)

  42. [50]

    Stenzel, R

    L. Stenzel, R. ter Hoeven, R. Miyazaki, T. Yamaji, M. Shirane, and W. Lechner, Phys. Rev. A113, 062621 (2026)

  43. [51]

    A. W. Glaetzle, R. M. W. van Bijnen, P. Zoller, and W. Lechner, Nature Communications8, 15813 (2017)

  44. [52]

    J. R. McClean, S. Boixo, V. N. Smelyanskiy, R. Babbush, and H. Neven, Nature Communications9, 4812 (2018)

  45. [53]

    Qaoa on frustrated ising rings,https://gitlab.com/AnitaW/frustrated-ising-rings

  46. [54]

    J. R. Johansson, P. D. Nation, and F. Nori, Computer Physics Communications183, 1760 (2012)

  47. [55]

    J. R. Johansson, P. D. Nation, and F. Nori, Computer Physics Communications184, 1234 (2013)

  48. [56]

    Virtanen, R

    P. Virtanen, R. Gommers, T. E. Oliphant,et al., Nature Methods17, 261 (2020)

  49. [57]

    Javadi-Abhari, M

    A. Javadi-Abhari, M. Treinish, K. Krsulich, C. J. Wood, J. Lishman, J. Gacon, S. Martiel, P. D. Nation, L. S. Bishop, A. W. Cross, B. R. Johnson, and J. M. Gambetta, Quantum computing with Qiskit (2024), arXiv:2405.08810 [quant-ph]

  50. [58]

    Nocedal and S

    J. Nocedal and S. J. Wright,Numerical optimization(Springer, 1999)

  51. [59]

    Albertini and D

    F. Albertini and D. D’Alessandro, Linear Algebra and its Applications350, 213 (2002)

  52. [60]

    Hence, one constraint is needed

    HereK= N, but one parity configuration can be decoded into two logical ones and, therefore, 2 N−1 physical states correspond to 2 N logical ones, the remainder are non-valid states. Hence, one constraint is needed

Pith tools

Reviewed August 16, 2026 · model on record in the stance chip above.