Pith. sign in

REVIEW 4 major objections 4 minor 47 references

State Stabilization for Gate-Model Quantum Computers

T0 review · 4 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read The paper claims that an orthogonal stabilizer matrix, found by minimizing temporal variation in gate parameters, can hold a gate-model quantum computer in its optimal state across arbitrarily many running sequences.

desk verdict Central theorem is asserted, not proven, and fails on a two-qubit example; the paper repackages slow feature analysis without a valid result. read the letter →

arxiv 1909.01044 v1 pith:TQZK4GFM submitted 2019-09-03 quant-ph cs.ITmath.IT

classification quant-phcs.ITmath.IT MSC 81P68 PACS 03.67.Lx
keywords quantumstatestabilizationgate-modelcomputersgateparameteroptimizationunsupervisedlearningstabilityclassesslowfeatureanalysisorthogonaltransformationnear-termcomputing
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

This paper tries to establish that the optimal state of a gate-model quantum computer can be stabilized across an arbitrary number of running sequences by an orthogonal transformation of the gate parameters. The authors define a stabilizer matrix $S$ that maps the sequence of optimal parameter vectors $\alpha$ to a stabilized sequence $\beta=S^T\alpha$, and they propose finding $S$ by minimizing the temporal variation between consecutive parameter vectors. They further claim that the stabilized state preserves the objective value of the optimal state, $f(\phi)=f(\theta^*)$, and that the stabilized sequences can be sorted into stability classes by unsupervised learning. A reader might care because holding a near-term quantum computer near its optimal operating point over many runs is a practical obstacle that this procedure addresses purely classically.

What carries the argument

The central object is the stabilizer matrix $S$, an orthogonal transformation that is supposed to convert the matrix of optimal gate parameters $\alpha$ into the matrix of stabilized parameters $\beta=S^T\alpha$ without changing the objective function value. The argument is carried by the objective $F^*$, which measures temporal variation: the term $\chi$ penalizes squared differences between neighboring stabilized parameter vectors, and the term $c\tau$ penalizes differences between more distant runs with weights that decay with the distance in the original parameter sequence. Minimizing $F^*$ is reduced to a generalized eigenvalue problem, and the eigenvectors supply $S$. The classification half of the paper is carried by the kernel machine $K$ and the quantities $\varphi_k(\phi_r)=(\nu_k(\phi_r))^T f^C_k(\phi_r)$, which compare stabilized sequences in a reproducing kernel Hilbert space and assign each sequence to a stability class.

What would settle it

Take a one-qubit circuit $U(\theta)=e^{-i\theta X}$ with objective $C=Z$ and two runs whose optimal parameters make the minimizer of $F^*$ nontrivial; compute the stabilized parameters $\phi=S^T\theta$, then compare $\langle\phi_r|Z|\phi_r\rangle$ with $\langle\theta_r|Z|\theta_r\rangle$ for each $r$. A single mismatch would show that the stabilized state is not optimal, contradicting the paper's central claim.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1: for $R$ running sequences, a stabilizer matrix $S$ with $S^TS=I$ can be found by minimizing $F^* = \arg\min_S \left(\operatorname{Tr}\left(S^T \Delta\alpha (\Delta\alpha)^T S\right) + c\tau\right)$, where $\Delta\alpha$ collects the differences between consecutive optimal parameter vectors, $\tau$ penalizes longer-range differences through weighted coefficients $\omega_{rs}$, and $c$ is a regularization constant. The minimization is rewritten as the generalized eigenvalue problem $\left(\Delta\alpha\,\sigma\,(\Delta\alpha)^T\right)S = \lambda\left(\Delta\alpha\,\eta\,(\Delta\alpha)^T\right)S$, and the solution yields $\beta=S^T\alpha$. The paper asserts that the resulting stabilized states $|\phi_r\rangle$ achieve $f(\phi)=f(\theta^*)$, i.e. the stabilization does not cost any objective-function value. Theorem 2 then assigns each stabilized sequence to one of $K$ stability classes through kernel-based correlation analysis in a high-dimensional Hilbert space, with the probabilistic classifier $f^C_k(\phi_{r,i})$ and weights $\nu_k(\phi_{r,i})=\phi_{r,i}/\pi$.

Load-bearing premise

The load-bearing premise is that changing the gate parameters from the optimal values in the particular way the stabilizer matrix prescribes never changes how well the problem is solved, even though the recipe for choosing that change never looks at the objective function or the cost operator.

Editorial extensions

If this is right

  • Any set of optimal parameter vectors $\alpha$ yields a stabilized set $\beta=S^T\alpha$ with reduced run-to-run temporal variation, because the objective $F^*$ is explicitly minimized for that purpose.
  • The stabilized state is claimed to preserve the objective value, $f(\phi)=f(\theta^*)$, so solution quality of the original problem is not degraded.
  • The stabilization is a classical computation on the gate parameters, so it can be applied to an existing circuit without altering the hardware unitaries.
  • For any number of runs $R$, each run receives a stability-class label, allowing drift to be monitored over time.
  • Because the learning procedures are unsupervised, no pre-labeled stable states are needed to find $S$ or the stability classes.

Reading between the lines

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

  • A direct extension, not tested in the paper, is to close the loop: recompute $S$ after each stabilization step and re-optimize the objective on the stabilized manifold, turning a one-shot correction into adaptive calibration.
  • The same machinery would transfer to variational and approximate-optimization parameter schedules, where drift between objective evaluations is a known obstacle, since the construction only uses the parameter sequences, not the specific circuit.
  • The invariance claim is the point most worth testing: a numerical sweep over random unitaries and cost operators would show whether $S$ must be restricted to transformations that commute with $C$, a stronger condition than orthogonality.
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 / 4 minor

Summary. The paper proposes a method to stabilize the optimal state of a gate-model quantum computer. Gate parameter vectors θ*_r for R running sequences are collected into a matrix α, and the stabilized parameters are obtained as β = Sᵀα, where S is claimed to be orthogonal. Theorem 1 asserts that S can be found by minimizing an objective function F* (Eq. 22) that penalizes temporal variation of the gate parameters. Theorem 2 proposes a kernel-based classification of the stabilized sequences into stability classes. Section 5 evaluates the relative entropy between stabilized and target parameters under an assumed sinusoidal model and computes a correlation coefficient between toy gate-parameter functions. The central claim is that the stabilized state |φ⟩ satisfies f(φ)=f(θ*), preserving optimality across all running sequences.

Significance. If the central claim were established, the idea of preserving optimality while stabilizing gate parameters through an orthogonal transformation would be a valuable contribution to near-term gate-model quantum computation, particularly combined with unsupervised learning. However, the manuscript does not prove the central invariance, and its numerical evaluation does not exercise the proposed algorithms. The paper provides no machine-checked proofs or reproducible code, and the only quantitative results are derived from an assumed sinusoidal model whose parameters are chosen ad hoc. Consequently, the work, as presented, does not establish any falsifiable prediction or validated method, and its significance is currently more suggestive than substantive.

major comments (4)
  1. [§3, Eq. (22)] The central claim of Theorem 1, that stabilizing via β=Sᵀα preserves f(φ)=f(θ*) as required by Eq. (9), is never proven and is in fact false in general. The objective function F* in Eq. (22) contains only the temporal-difference term χ and the weighted term cτ; neither term involves the cost operator C or the objective function f. An orthogonal linear combination of optimal parameter vectors need not be optimal. A concrete counterexample is given by L=2, U_i(θ)=exp(-iθY), input |00>, and C=|++⟩⟨++| with |+⟩=(|0⟩+|1⟩)/√2. Then f(θ)=((1+sin θ1)(1+sin θ2))/4, whose unique maximum on [0,π]² is θ*=(π/2,π/2). Taking α columns θ*_1=(π/2,π/2) and θ*_2=(π/2,π/2+δ), the 45-degree rotation S gives β_1=Sᵀθ*_1=(π/√2,0), and f(β_1)=(1+sin(π/√2))/4≈0.45, not 1. Since F* depends only on ∆α and the weights ω_rs, this S and the identity are indistinguishable to F*, so the minimization cannot rule out a stabilizer that destroys optimality.
  2. [§3, Eqs. (24), (28)] The S matrix obtained from the generalized eigenvalue problem in Eq. (28) is not shown to satisfy the orthogonality condition SᵀS=I imposed in Eq. (13). The constraint that is actually imposed is (∆β)ᵀη∆β=I, i.e., Sᵀ(∆αη∆αᵀ)S=I, and η itself depends on α through the weight coefficients (20). Even if a solution to Eq. (28) exists, it is an orthogonal transformation with respect to the matrix ∆αη∆αᵀ, not a Euclidean orthogonal matrix, and therefore β=Sᵀα does not necessarily satisfy the stated stabilization relation. The proof of Theorem 1 needs to establish that the eigenvector matrix can be chosen so that SᵀS=I, and this is not done.
  3. [§5, Eqs. (60)–(64)] The numerical evaluation in Section 5 does not implement or simulate Algorithm 1 (AS) or Algorithm 2 (AC). Instead, the stability parameter δ(r) is derived entirely from the assumed sinusoidal model (60) for f_D(β‖β*), with parameters γ, λ, N, c, and E(D) chosen ad hoc. Equation (64) then yields δ(r)=1/N, which is an inverse of the number of oscillations of the assumed sinusoidal function; it is a property of that assumed model, not of the proposed stabilization algorithm. The correlation coefficient µ(β,β*) in §5.2 is computed for arbitrary functions f(φ_r) defined in Eqs. (67) and (69) and is never connected to the output β of Algorithm 1. Consequently, Section 5 provides no numerical evidence that the proposed method achieves state stabilization or preserves optimality.
  4. [§4, Algorithm 2, Step 7] The classification procedure in Theorem 2 and Algorithm 2 is not operationally defined. The classifier f_C^k in Eq. (39) is asserted to output probabilities, but no training procedure, loss function, or connection to the stabilized data β from Algorithm 1 is given. Moreover, Step 7 defines C(φ_r)=ξ(φ_r)C_p+ℓ_k(φ_r)C_q, but the addition of stability classes C_p and C_q, and multiplication by scalar parameters, is not defined. It is unclear whether the output is a class label, a vector, or a fuzzy membership vector. This makes the classification claim unfalsifiable as stated.
minor comments (4)
  1. [§2, Eq. (3)] The text says 'where C is the classical value of the objective function', but Eq. (3) uses C as an operator in the expectation value ⟨θ*|C|θ*⟩. This is inconsistent and should be clarified.
  2. [§3, Eq. (20)] The weight coefficient ω_rs depends on the squared norm of differences of ∆θ* vectors, but the manuscript does not specify the norm or explain how the kernel width ζ is chosen. The same symbol ζ is also used later; please disambiguate.
  3. [§5, Eq. (60)] The constant c in Eq. (60) is reused for the amplitude of the sinusoidal model, while c was already used as the regularization constant in Eq. (22) and as the kernel bandwidth in Eq. (47). This triplication of the symbol c is confusing.
  4. [Throughout] There are several typographical errors, including 'uniaries' and 'quatum', and some equations have inconsistent bracket formatting (e.g., Eq. (2) and Eq. (7)). These should be corrected.

Circularity Check

2 steps flagged · score 6.0 of 10

Eq. (9) is never derived: minimizing F* in Eq. (22) contains no cost operator C, so the claim that stabilized β preserves f(θ*) is an unproven redefinition; the δ=1/N 'stability' result is likewise fixed by the assumed sinusoidal ansatz.

  1. self definitional [Theorem 1, Section 3, Eqs. (9) and (22)]
    "such that the objective function value is stabilized into f(ϕ)=⟨ϕ|C|ϕ⟩=f(θ∗) ... At a particular S in equations (18) and (21), the stabilization of the optimal state of the quantum computer through the R sequences can be reformulated via an objective function F∗, subject to a minimization as F∗=arg min_S(χ+cτ)."

    Equation (9) defines the stabilized state by f(φ)=f(θ*), i.e., by preserving the objective value. Equation (22) defines 'the stabilization of the optimal state' as the minimizer of F*=χ+cτ, with χ=Tr(S^T Δα Δα^T S) and τ built from γ_rs=ω_rs‖Δφ_r−Δφ_s‖². Neither F* nor the accompanying generalized eigenproblem (28) contains the cost operator C or the function f, so the S delivered by the minimization is only a minimizer of temporal gate-parameter variation. The theorem's conclusion that this S yields β with f(φ)=f(θ*) is therefore not derived; it holds only if 'stabilization' is redefined to be F*-minimization. Orthogonal mixtures of optimal parameter vectors need not be optimal, so Eq. (9) does not follow by construction.

  2. fitted input called prediction [Section 5.1, Eqs. (57)-(64)]
    "For analytical purposes, let us assume that fD(β‖β∗)(r) oscillates between a minimal value γ≥0, and a maximal value γ≤λ≤1 ... therefore fD(β‖β∗)(r) can be rewritten as fD(β‖β∗)(r)=c sin(N 2π r/R)+E(D(β‖β∗)). ... then, by using (63), the quantity in (57) is as δ(r)=(R/(2π)√(N^2 4π^2/R^2))^{-1}=1/N, that identifies the inverse of the number of oscillations."

    The stability parameter δ(r) is defined in Eq. (57) from Δ(f_D), the integrated second derivative of the relative-entropy curve. To evaluate it, Eq. (60) assumes f_D is a sinusoid with a free parameter N ('number of oscillations'), with c and E fixed as (λ−γ)/2 and c+γ. Under this assumed form the second derivative integrates to N²4π²/R², so δ=1/N is an algebraic identity. The 'stability' of the system state is therefore just the reciprocal of an arbitrarily chosen input N; no property of Algorithm 1 or of β enters. A free parameter is renamed as a stability result.

full rationale

The central theorem is not supported by a derivation: the paper defines the stabilized state by f(φ)=f(θ*) in Eq. (9), but the S chosen by Eqs. (22)/(28) minimizes only a temporal-variation functional of Δα and weight coefficients, with no term involving C or f. Since an orthogonal map β=S^T α can send optimal parameter vectors into non-optimal ones (e.g., L=2, θ*=(π/2,π/2), a 45-degree rotation gives f≈0.45 rather than 1), Eq. (9) does not follow from the minimization; the theorem is true only if 'stabilization' is redefined to mean F*-minimization. The numerical stability parameter is also a tautology: Eq. (57) defines δ from the second derivative of the relative-entropy curve, and Eq. (60) assumes a sinusoid with free oscillation count N, making δ=1/N an algebraic identity. No external benchmark, code, or independent theorem is supplied, and the generalized eigenproblem (28) is not shown to yield an orthogonal S (it imposes S^T Δα η Δα^T S=I, not S^T S=I). These are load-bearing gaps in the paper's own equations; however, the paper does not rely on a self-citation chain, so the circularity is partial rather than complete.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central formal machinery rests on several unstated choices: the availability of optimal gate parameters, the preservation of the objective function under an orthogonal transformation (never proven), and a sinusoidal approximation for the relative entropy. These are ad hoc within the paper and not backed by prior literature or independent evidence.

free parameters (5)
  • Regularization constant c = not specified
    Introduced in eq. (22) to balance chi and tau; no method for choosing it is given.
  • Kernel width kappa and scale zeta = not specified
    Parameters in the weight definition (20); no values or selection procedure provided.
  • Number of stability classes K = not specified
    Set of classes in (37); the paper does not explain how to choose K.
  • Gaussian kernel width (c=2 sigma^2) = not specified
    Used in eq. (47) for the classification kernel; not specified.
  • Numerical example parameters (N, C, C*, E(D)) = various, e.g., N=1, C=0.125, C*=0.1
    Parameters chosen for illustration in Section 5; they are not fitted to data but are free variables that determine the plots.
assumptions (4)
  • domain assumption The optimal gate parameter vectors theta*_r are known for each running sequence r.
    The method requires alpha (eq. 10) as input, but the paper does not address how these optimal parameters are obtained; finding them is generally a hard optimization problem.
  • ad hoc to paper The orthogonal transformation beta = S^T alpha preserves the objective function value f(phi) = f(theta*).
    Stated in eq. (9) as a requirement, but never derived or enforced in the optimization of F* (eq. 22); this is the load-bearing premise that is not justified.
  • ad hoc to paper The relative entropy function f_D(beta||beta*) can be approximated as a sinusoidal function (60) with constant expected value for analytical tractability.
    Assumed in Section 5.1 to compute the stability parameter delta; no justification that this approximates any actual quantum system.
  • ad hoc to paper The constraint (Delta beta)^T eta Delta beta = I can be imposed without loss of generality.
    Equation (24) is set as a normalization constraint without physical motivation; it is characteristic of SFA but not justified in the quantum context.

how reviews work

0 comments
Cite this review

Pith. "Pith review of State Stabilization for Gate-Model Quantum Computers." pith.science (2026). https://pith.science/paper/TQZK4GFM

@misc{pith2026190901044,
  author       = {Pith},
  title        = {Pith review of: State Stabilization for Gate-Model Quantum Computers},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TQZK4GFM}},
  note         = {Machine review of arXiv:1909.01044}
}
read the original abstract

Gate-model quantum computers can allow quantum computations in near-term implementations. The stabilization of an optimal quantum state of a quantum computer is a challenge, since it requires stable quantum evolutions via a precise calibration of the unitaries. Here, we propose a method for the stabilization of an optimal quantum state of a quantum computer through an arbitrary number of running sequences. The optimal state of the quantum computer is set to maximize an objective function of an arbitrary problem fed into the quantum computer. We also propose a procedure to classify the stabilized quantum states of the quantum computer into stability classes. The results are convenient for gate-model quantum computations and near-term quantum computers.

Figures

Figures reproduced from arXiv: 1909.01044 by the authors.

Figure 1
Figure 1. The framework F for the stabilization of the optimal state of the quantum computer and the stability-class determination. In the R running sequences, R input systems are fed into the input of the quantum computer, in an r-th running sequence, r = 1, . . . , R, an r-th input system, |ψri = P i αi |ii, is evolved via the sequence of the L uniaries of the quatum computer. The R running sequences identify an input syste… view at source ↗
Figure 2
Figure 2. The fD( βkβ∗) (r) relative entropy values between the gate parameters of the stabilized system |~ϕri and target system |~ϕ ∗ r i for R running sequences, r = 1, . . . R, R = 10, E (D (βk β ∗ )) = 0.1. (a) N = 1. (b) N = 2. (c) N = 3. (d). Stability parameter δ (r) for the different relative entropy values. 13 [PITH_FULL_IMAGE:figures/full_fig_p013_2.png] view at source ↗
Figure 3
Figure 3. Gate parameter values of f (~ϕr) and f (~ϕ ∗ r ) in function of running sequence r, r = 1, . . . , R, R = 10, L = 1, for different N, C and C ∗ , C ∈ [0, 1], C ∗ ∈ [0, 1]. (a-b) N = 1, C = 0.125, C ∗ = 0.1. (c-d) N = 1, C = 0.3, C ∗ = 0.2 (e-f) N = 2, C = 0.5, C ∗ = 0.3. (g) The µ (β, β∗ ) correlation values between the gate parameters of (a-f), S is an indexing parameter. 15 [PITH_FULL_IMAGE:figures/full_fig_p015_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: The distribution of µ (β, β∗ ) in function of C and C ∗ at f (~ϕr) and f (~ϕ ∗ r ), L = 1, R = 10. (a) N = 1. (b) N = 2. (c) N = 3. 6 Conclusions Here, we defined a method for the learning of stable quantum evolutions in gate-model quantum computer architectures. The m…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

47 extracted references · 17 canonical work pages

  1. [1]

    Quantum Computing in the NISQ era and beyond, Quantum 2, 79 (2018)

    Preskill, J. Quantum Computing in the NISQ era and beyond, Quantum 2, 79 (2018)

  2. [2]

    Harrow, A. W. and Montanaro, A. Quantum Computational Supremacy, Nature, vol 549, pages 203-209 (2017). 16

  3. [3]

    and Chen, L

    Aaronson, S. and Chen, L. Complexity-theoretic foundations of quantum supremacy experi- ments. Proceedings of the 32nd Computational Complexity Conference , CCC ’17, pages 22:1- 22:67, (2017)

  4. [4]

    Biamonte, J. et al. Quantum Machine Learning. Nature, 549, 195-202 (2017)

  5. [5]

    and Hinton, G

    LeCun, Y., Bengio, Y. and Hinton, G. Deep Learning. Nature 521, 436-444 (2014)

  6. [6]

    and Courville, A

    Goodfellow, I., Bengio, Y. and Courville, A. Deep Learning. MIT Press. Cambridge, MA, 2016

  7. [7]

    Debnath, S. et al. Demonstration of a small programmable quantum computer with atomic qubits. Nature 536, 63-66 (2016)

  8. [8]

    Monz, T. et al. Realization of a scalable Shor algorithm. Science 351, 1068-1070 (2016)

Show all 47 references
  1. [9]

    Barends, R. et al. Superconducting quantum circuits at the surface code threshold for fault tolerance. Nature 508, 500-503 (2014)

  2. [10]

    and Wineland, D

    Kielpinski, D., Monroe, C. and Wineland, D. J. Architecture for a large-scale ion-trap quantum computer. Nature 417, 709-711 (2002)

  3. [11]

    Ofek, N. et al. Extending the lifetime of a quantum bit with error correction in superconducting circuits. Nature 536, 441-445 (2016)

  4. [12]

    and Gutmann, S

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

  5. [13]

    and Neven, H

    Farhi, E., Goldstone, J., Gutmann, S. and Neven, H. Quantum Algorithms for Fixed Qubit Architectures. arXiv:1703.06199v1 (2017)

  6. [14]

    and Neven, H

    Farhi, E. and Neven, H. Classification with Quantum Neural Networks on Near Term Proces- sors, arXiv:1802.06002v1 (2018)

  7. [15]

    and Gutmann, S

    Farhi, E., Goldstone, J. and Gutmann, S. A Quantum Approximate Optimization Algorithm Applied to a Bounded Occurrence Constraint Problem. arXiv:1412.6062. (2014)

  8. [16]

    and Lloyd, S

    Rebentrost, P., Mohseni, M. and Lloyd, S. Quantum Support Vector Machine for Big Data Classification. Phys. Rev. Lett. 113. (2014)

  9. [17]

    The Universe as Quantum Computer, A Computable Universe: Understand- ing and exploring Nature as computation , H

    Lloyd, S. The Universe as Quantum Computer, A Computable Universe: Understand- ing and exploring Nature as computation , H. Zenil ed., World Scientific, Singapore, 2012, arXiv:1312.4455v1 (2013)

  10. [18]

    and Rebentrost, P

    Lloyd, S., Mohseni, M. and Rebentrost, P. Quantum algorithms for supervised and unsuper- vised machine learning, arXiv:1307.0411v2 (2013)

  11. [19]

    and Zanardi, P

    Lloyd, S., Garnerone, S. and Zanardi, P. Quantum algorithms for topological and geometric analysis of data. Nat. Commun., 7, arXiv:1408.3106 (2016)

  12. [20]

    H., Wong, F

    Lloyd, S., Shapiro, J. H., Wong, F. N. C., Kumar, P., Shahriar, S. M. and Yuen, H. P. Infrastructure for the quantum Internet. ACM SIGCOMM Computer Communication Review, 34, 9-20 (2004). 17

  13. [21]

    and Rebentrost, P

    Lloyd, S., Mohseni, M. and Rebentrost, P. Quantum principal component analysis. Nature Physics, 10, 631 (2014)

  14. [22]

    and Nguyen, H

    Gyongyosi, L., Imre, S. and Nguyen, H. V. A Survey on Quantum Channel Capacities, IEEE Communications Surveys and Tutorials 99, 1, doi: 10.1109/COMST.2017.2786748 (2018)

  15. [23]

    and Petruccione, F

    Schuld, M., Sinayskiy, I. and Petruccione, F. An introduction to quantum machine learning. Contemporary Physics 56, pp. 172-185. arXiv: 1409.3097 (2015)

  16. [24]

    Quantum Networking , John Wiley and Sons Ltd, ISBN 1118648927, 9781118648926 (2014)

    Van Meter, R. Quantum Networking , John Wiley and Sons Ltd, ISBN 1118648927, 9781118648926 (2014)

  17. [25]

    and Gyongyosi, L

    Imre, S. and Gyongyosi, L. Advanced Quantum Communications - An Engineering Approach . Wiley-IEEE Press (New Jersey, USA), (2012)

  18. [26]

    and Banchi, L

    Pirandola, S., Laurenza, R., Ottaviani, C. and Banchi, L. Fundamental limits of repeaterless quantum communications, Nature Communications, 15043, doi:10.1038/ncomms15043 (2017)

  19. [27]

    L., Laurenza, R., Ottaviani, C., Cope, T

    Pirandola, S., Braunstein, S. L., Laurenza, R., Ottaviani, C., Cope, T. P. W., Spedalieri, G. and Banchi, L. Theory of channel simulation and bounds for private communication, Quantum Sci. Technol. 3, 035009 (2018)

  20. [28]

    Capacities of repeater-assisted quantum communications, arXiv:1601.00966 (2016)

    Pirandola, S. Capacities of repeater-assisted quantum communications, arXiv:1601.00966 (2016)

  21. [29]

    End-to-end capacities of a quantum communication network, Commun

    Pirandola, S. End-to-end capacities of a quantum communication network, Commun. Phys. 2, 51 (2019)

  22. [30]

    Quantum Information Theory and Quantum Statistics , Springer-Verlag, Heidelberg, Hiv: 6

    Petz, D. Quantum Information Theory and Quantum Statistics , Springer-Verlag, Heidelberg, Hiv: 6. (2008)

  23. [31]

    S., Elons, A

    Chawky, B. S., Elons, A. S., Ali, A. and Shedeed, H. A. A Study of Action Recognition Problems: Dataset and Architectures Perspectives, In: Hassanien, A. E. and Oliva, D. A. (eds.), Advances in Soft Computing and Machine Learning in Image Processing , Studies in Computational ...

  24. [32]

    and Tao, D

    Miao, J., Xu, X., Xing, X. and Tao, D. Manifold Regularized Slow Feature Analysis for Dynamic Texture Recognition, arXiv:1706.03015v1 (2017)

  25. [33]

    and Sejnowski, T

    Wiskott, L. and Sejnowski, T. J. Slow feature analysis: Unsupervised learning of invariances, Neural Computation, vol. 14, pp. 715–770, (2002)

  26. [34]

    R., Scholz, M

    Mika, S., Scholkopf, B., Smola, A., Muller, K. R., Scholz, M. and Ratsch, G. Kernel pca and de-noising in feature spaces, Advances in Neural Information Processing Systems 11. pp. 536–542, MIT Press, (1999)

  27. [35]

    and Cristianini, N

    Shawe-Taylor, J. and Cristianini, N. Kernel Methods for Pattern Analysis . Cambridge Univer- sity Press (2004)

  28. [36]

    and Haykin, S

    Liu, W., Principe, J. and Haykin, S. Kernel Adaptive Filtering: A Comprehensive Introduction. Wiley (2010). 18

  29. [37]

    and Gould, S

    Cherian, A. and Gould, S. Second-order Temporal Pooling for Action Recognition, arXiv:1704.06925v1 (2017)

  30. [38]

    Zhou, X., Leung, D. W. and Chuang, I. L. Methodology for quantum logic gate construction, Phys. Rev. A , vol. 62, p. 052316 (2000)

  31. [39]

    Gottesman, D., Chuang, I. L. Quantum Teleportation is a Universal Computational Primitive, Nature 402, 390-393 (1999)

  32. [40]

    and Roetteler, M

    Amy, M., Maslov, D., Mosca, M. and Roetteler, M. A meet-in-the middle algorithm for fast synthesis of depth-optimal quantum circuits, IEEE Transactions on Computer-Aided Design of Integrated Circuits and Systems , vol. 32, no. 6, pp. 818–830, (2013)

  33. [41]

    and Devitt, S

    Paler, A., Polian, I., Nemoto, K. and Devitt, S. J. Fault-tolerant, high level quantum circuits: form, compilation and description, Quantum Science and Technology, vol. 2, no. 2, p. 025003, (2017)

  34. [42]

    Brandao, F. G. S. L., Broughton, M., Farhi, E., Gutmann, S. and Neven, H. For Fixed Control Parameters the Quantum Approximate Optimization Algorithm’s Objective Function Value Concentrates for Typical Instances, arXiv:1812.04170 (2018)

  35. [43]

    and Lukin, M

    Zhou, L., Wang, S.-T., Choi, S., Pichler, H. and Lukin, M. D. Quantum Approximate Opti- mization Algorithm: Performance, Mechanism, and Implementation on Near-Term Devices, arXiv:1812.01041 (2018)

  36. [44]

    Quantum Approximate Optimization with Parallelizable Gates, arXiv:1802.01157v2 (2018)

    Lechner, W. Quantum Approximate Optimization with Parallelizable Gates, arXiv:1802.01157v2 (2018)

  37. [45]

    Crooks, G. E. Performance of the Quantum Approximate Optimization Algorithm on the Maximum Cut Problem, arXiv:1811.08419 (2018)

  38. [46]

    W., Jonay, C

    Ho, W. W., Jonay, C. and Hsieh, T. H. Ultrafast State Preparation via the Quantum Approx- imate Optimization Algorithm with Long Range Interactions, arXiv:1810.04817 (2018)

  39. [47]

    Song, C. et al. 10-Qubit Entanglement and Parallel Logic Operations with a Superconducting Circuit, Physical Review Letters, vol. 119, no. 18, p. 180511 (2017). 19 A Appendix A.1 Abbreviations QG Quantum Gate structure of a gate-model quantum computer RKHS Reproducing Kernel H...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.