Pith. sign in

REVIEW 3 major objections 6 minor 75 references

Hamiltonian Expressibility for Ansatz Selection in Variational Quantum Algorithms

T0 review · 3 major / 6 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read Hamiltonian expressibility predicts VQE answer quality in 4-qubit problems, but the optimal level flips with the ground-state structure: low expressibility for basis states, high expressibility for superpositions, and intermediate…

desk verdict Useful empirical study of a plausible heuristic, but the central claim that expressibility (rather than depth) drives ansatz performance is not established. read the letter →

arxiv 2507.22550 v2 pith:OTZ6IFBT submitted 2025-07-30 quant-ph cs.ET

Filippo Brozzi , Gloria Turati , Maurizio Ferrari Dacrema , Filippo Caruso , Paolo Cremonesi This is my paper
classification quant-phcs.ET MSC 81P68
keywords variationalquantumalgorithmsHamiltonianexpressibilityansatzselectioneigensolverbarrenplateauscircuitdepthnoiseresilienceframepotentials
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 Hamiltonian expressibility, a circuit's ability to uniformly explore the energy landscape of a specific Hamiltonian, can guide ansatz selection in variational quantum eigensolvers, at least for small systems. Working with 19 circuit families at 4 and 8 qubits, it estimates expressibility by Monte Carlo sampling of frame potentials, trains each circuit with VQE on families of diagonal and non-diagonal Hamiltonians, and correlates expressibility with the normalized approximation ratio. The central result is a problem-dependent recommendation: for problems whose ground states are basis states, meaning classical binary strings, low-expressibility circuits are preferable, while for problems with superposition ground states from non-diagonal Hamiltonians, high-expressibility circuits perform better under ideal or low-noise conditions. Under stronger noise, low expressibility is still preferred for basis-state problems, and intermediate expressibility becomes best for some superposition-state problems. The correlation weakens at 8 qubits, which the paper interprets as the expected onset of barren-plateau trainability limits.

What carries the argument

The central object is the Hamiltonian expressibility $\varepsilon_H(U,H)$, defined through the frame-potential difference $F(U,H)-F_{\mathrm{Haar}}(H)$, where $F(U,H)$ is the ansatz-Hamiltonian frame potential averaged over parameter pairs and $F_{\mathrm{Haar}}(H)$ is the closed-form Haar value. Smaller $\varepsilon_H$, or a ratio $\gamma_H=F/F_{\mathrm{Haar}}$ closer to 1, means the circuit ensemble samples the energy landscape more like the Haar-uniform distribution. The paper estimates $F(U,H)$ with a Monte Carlo trace sampling procedure, uses Haar-sampled unitaries to set finite-sample thresholds for maximal expressibility, and then compares these metrics, via Pearson, Spearman, Kendall tau, and mutual information, to the average normalized approximation ratio of VQE runs.

What would settle it

Run a depth-controlled VQE study: fix the number of layers, gate count, and parameter count across a set of 4-qubit ansatze on the same Hamiltonians, choosing architectures whose Hamiltonian expressibility differs, for example by reordering or replacing entangling gates; if approximation ratios do not track expressibility within each fixed-depth group, the paper's central claim is refuted.

Watch

Extended reading notes

Core claim

On its own terms, the paper claims that Hamiltonian expressibility is a meaningful, problem-dependent predictor of VQE solution quality, with a direction that flips with the structure of the ground state. For 4-qubit noiseless VQE runs, Spearman and Kendall correlation coefficients between expressibility and approximation ratio are strongly negative for Heisenberg superposition-state and random non-diagonal problems, meaning better expressibility tracks better accuracy, while diagonal and basis-state problems show weak or positive monotonic correlations, meaning low-expressibility circuits are sufficient or even preferable. The paper further claims that adding layers increases Hamiltonian expressibility until saturation, that all tested circuits become more expressive on diagonal Hamiltonians, and that the predictive power weakens as qubit count grows, consistent with high expressibility being linked to barren plateaus. Under simulated hardware noise, the clear ideal-case dichotomy partly collapses: basis-state problems still favor low expressibility, but superposition-state problems are best served by intermediate expressibility, because deep, highly expressive circuits accumulate more noise.

Load-bearing premise

The load-bearing premise is that the observed correlation between Hamiltonian expressibility and solution quality is causal and specific to expressibility; because the 4-qubit circuits range from 1 to 5 layers and expressibility grows with depth, the results could instead be explained by circuit depth, parameter count, or noise sensitivity, which would invalidate the problem-dependent recommendations.

Editorial extensions

If this is right

  • For small-scale diagonal QUBO problems, practitioners can deliberately choose shallow, low-expressibility circuits; the paper finds these yield better approximation ratios than highly expressive ones.
  • For non-diagonal problems with superposition ground states under ideal or low-noise conditions, selecting high-expressibility circuits improves VQE solution quality.
  • Expressibility-based screening can be done before training: the Monte Carlo frame-potential estimate is cheaper than repeated VQE runs, so candidate ansatze can be ranked by expected fit to the problem class.
  • As system size grows, expressibility becomes a weaker guide because high expressibility is tied to barren-plateau trainability loss; the paper expects low-expressibility circuits to remain safer for basis-state problems, while high-expressibility advantages fade for superposition problems.

Reading between the lines

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

  • Because the 4-qubit circuits vary in depth from 1 to 5 layers and the paper itself shows expressibility saturates with depth, the cleanest reading is that depth, not expressibility alone, may be doing much of the work; a same-depth comparison across architectures would settle this.
  • The bell-shaped dependence seen under noise suggests a testable optimal expressibility window for each problem class; one could sweep noise parameters and track where the peak approximation ratio sits.
  • The monotonic saturation of expressibility with layers could be turned into a stopping criterion for adaptive ansatz construction: stop adding layers once the circuit crosses the Haar threshold, since extra depth adds noise without expanding the explored energy landscape.
  • The 8-qubit result being limited to single-layer circuits means the weakened correlations may reflect the depth truncation as much as barren plateaus; replicating the 4-qubit depth sweep at 8 qubits would separate the two explanations.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. This paper studies Hamiltonian expressibility as a criterion for ansatz selection in the Variational Quantum Eigensolver (VQE). The authors estimate the frame-potential-based metrics ε_H and γ_H for 95 four-qubit circuits (19 architectures with 1–5 layers) and 19 one-layer eight-qubit circuits on a battery of diagonal and non-diagonal Hamiltonians, using a Monte Carlo procedure with k = 250,000 samples and Haar-derived thresholds. They run noiseless and noisy VQE simulations, score each ansatz by its average normalized approximation ratio, and compute Pearson, Spearman, Kendall tau, and mutual information between expressibility and solution quality. Based on the correlations, they conclude that high Hamiltonian expressibility is beneficial for non-diagonal Hamiltonians with superposition ground states under ideal or low-noise conditions, that low expressibility is preferable for diagonal or basis-state problems, and that under stronger noise, intermediate expressibility becomes optimal for some superposition-state problems.

Significance. If the central claim held, the paper would provide a practical, pre-computable heuristic for choosing among candidate ansätze in small-scale VQAs and would sharpen the known expressibility–trainability trade-off with problem-dependent nuance. The manuscript has genuine strengths: the expressibility metrics are defined independently of the VQE outcomes; the Monte Carlo estimator is carefully calibrated (large sample size, Haar-derived thresholds, confidence intervals); the circuit and Hamiltonian sets are broad; and multiple correlation measures are reported. However, the empirical inference is currently confounded with circuit depth, so the stated problem-dependent recommendations are not yet established.

major comments (3)
  1. [§3.1.1, §4.1.1, §3.4.2] The paper's central claim is not yet supported because depth and expressibility co-vary in the pooled analysis. Section 3.1.1 constructs 95 four-qubit circuits by varying 19 base architectures over 1–5 layers, and Section 4.1.1 shows that Hamiltonian expressibility improves with depth. The correlation analysis in Section 3.4.2 (Figures 5, 6, and 7) pools all 95 circuits, so the observed negative correlations for non-diagonal/superposition classes and the positive or weak correlations for diagonal/basis classes could be fully mediated by depth, parameter count, or the noise sensitivity of deeper circuits. The 8-qubit experiments, which fix depth at one layer, show weakened correlations (Figure 8), exactly what the depth-mediated alternative predicts. The authors should add depth-controlled evidence: for example, partial correlations controlling for layer count, stratified correlations within each depth, or pairwise comparisons of circuits with the same depth but different expressibility. Without such evidence, the recommendation to select by expressibility rather than by depth is not established.
  2. [§4.2.1, Figures 5, 8, 10] The correlation claims lack significance assessment. The figures report mean correlation coefficients with standard deviations across problem instances, but no p-values, confidence intervals, or multiple-comparison corrections are given, and several reported coefficients are small (for example, the near-zero Pearson values for MinVertex and Heisenberg Basis State in Figure 5). The qualitative conclusions, such as 'low expressibility is preferable' or 'intermediate expressibility is optimal', rely on discriminating these small coefficients from zero. The authors should report significance tests or bootstrap confidence intervals for the reported correlations, or explicitly restrict their claims to coefficients exceeding a pre-specified threshold.
  3. [§4.2.2, Figures 12 and 14] The claim that intermediate Hamiltonian expressibility is optimal under noise for some superposition-state problems is supported only by visual inspection of two scatter plots for selected Heisenberg instances. The paper does not quantify the location or width of the 'optimal' expressibility range, does not test whether the peak is statistically distinguishable from a monotone trend, and does not check whether the intermediate-expressibility circuits are simply intermediate-depth circuits. This conclusion should either be reworded as a qualitative observation or supported by a fitted model or a formal test that also controls for depth.
minor comments (6)
  1. [§3.1.2] The classification of Hamiltonians into basis-state vs. superposition-state is performed after computing the ground states; the authors disclose this, but the subsequent correlation analysis treats these as fixed problem categories. A sentence acknowledging that the classes are data-dependent and that the split may inflate the apparent contrast would help the reader calibrate the results.
  2. [Figures 6 and 7] The colors encode the base circuit pattern, but no legend is provided, so the reader cannot identify which points correspond to which architecture or layer count. Adding a legend or point labels would improve interpretability.
  3. [Table 1] The table contains reference markers such as 'schedule1' whose corresponding footnotes do not appear in the extracted text; the authors should add the notes or remove the markers.
  4. [§3.3.1] The hardware topology is described only as 'inspired by commercially available processors'. Specifying the coupling map and gate set used in the simulations would improve reproducibility.
  5. [Appendix C] The estimators (C7) and (C8) use a threshold that is itself estimated from Haar sampling, but the uncertainty of this threshold is not propagated into the final expressibility values. A brief discussion of this limitation would be useful.
  6. [General] No code or data availability statement is included. For a benchmarking study of this kind, providing a public repository or explicit availability statement would strengthen reproducibility.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: Hamiltonian expressibility is estimated independently of VQE outcomes; the expressibility–performance link is an empirical correlation, not a construction.

full rationale

The derivation chain is not circular. The Hamiltonian expressibility metric ε_H (Eq. 8) and ratio γ_H (Eq. 10) are defined from frame potentials (Eqs. 3–6) that involve only the ansatz ensemble, the Haar distribution, and the Hamiltonian; they do not involve the VQE objective, the approximation ratio, or any fitted VQE parameter. The Monte Carlo estimator in Section 3.2 samples random parameter sets and Haar unitaries, and the thresholds in Appendix C are benchmarks obtained by sampling from the Haar distribution. VQE is executed independently in Section 3.3, and the normalized approximation ratio (Eq. 12) is computed from those runs. The correlation analysis in Section 3.4 is an empirical comparison of independently estimated quantities, with no parameter fitted from A.R. and then renamed as a prediction. The central claim that high or low expressibility correlates with performance is not derived from the definition of ε_H alone. The main validity caveat is that the 4-qubit circuit set varies depth from 1 to 5 and Section 4.1.1 shows expressibility improves with depth, so the pooled correlations in Figures 5–10 may be confounded by depth; however, this is an identification/confound concern, not circularity, and the 8-qubit fixed-depth analysis and per-class trend plots partially address it. Self-citations (Turati et al. 2023; Foderà et al. 2024) appear only in background lists of adaptive and reinforcement-learning methods and are not load-bearing. The barren-plateau interpretation cites the external Holmes et al. (2022). Hence no step of the paper reduces to its own inputs by construction.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central claim depends on the expressibility metric (predefined in prior work), the Monte Carlo sample size, the VQE simulation and noise model, and the a posteriori problem classification. No new physical entities are introduced.

free parameters (3)
  • Monte Carlo sample size k = 250,000
    Chosen to achieve ~0.5% relative accuracy on frame potential estimates (Appendix B); it affects the precision of the expressibility values that are the independent variable in all correlations.
  • QUBO penalty p = 8
    Used in MinVertexCover and MaxClique Hamiltonians (Eq. A2, A3); chosen to enforce constraints, a standard choice but impacts the Hamiltonians.
  • Noise parameters (T1, T2, err1, err2) = T1=T2=200us, err1=1.6e-4, err2=4e-3
    Set to model a generic backend; variations of these parameters are used for the noise-sweep in Figure 13. They influence the noisy VQE results and thus the correlations.
assumptions (4)
  • standard math The closed-form Haar-Hamiltonian frame potential F_Haar(H) (Eq. 6) from Holmes et al. (2022) is correct.
    Used to compute both expressibility metrics; correctness is assumed from prior work.
  • domain assumption Uniform sampling over the ansatz parameter space Theta yields a representative measure dW for the circuit ensemble C(U).
    Inherent to all expressibility estimates (Section 2.2.1); if parameter sampling does not represent the reachable unitaries, the metric may be biased.
  • domain assumption Exact <H>min and <H>max are known for each problem instance to compute the approximation ratio.
    Required by Eq. (12); the paper uses exact diagonalization for benchmarking, which is only possible in simulation.
  • domain assumption The Qiskit GenericBackendV2 noise model with T1=T2 and err2=25*err1 approximates real hardware behavior.
    Used in the noisy setting (Section 3.3.2); the specific relationship may not hold for all real devices.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Hamiltonian Expressibility for Ansatz Selection in Variational Quantum Algorithms." pith.science (2026). https://pith.science/paper/OTZ6IFBT

@misc{pith2026250722550,
  author       = {Pith},
  title        = {Pith review of: Hamiltonian Expressibility for Ansatz Selection in Variational Quantum Algorithms},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OTZ6IFBT}},
  note         = {Machine review of arXiv:2507.22550}
}
read the original abstract

In the context of Variational Quantum Algorithms (VQAs), selecting an appropriate ansatz is crucial for efficient problem-solving. Hamiltonian expressibility has been introduced as a metric to quantify a circuit's ability to uniformly explore the energy landscape associated with a Hamiltonian ground state search problem. However, its influence on solution quality remains largely unexplored. In this work, we estimate the Hamiltonian expressibility of a well-defined set of circuits applied to various Hamiltonians using a Monte Carlo-based approach. We analyze how ansatz depth influences expressibility and identify the most and least expressive circuits across different problem types. We then train each ansatz using the Variational Quantum Eigensolver (VQE) and analyze the correlation between solution quality and expressibility.Our results indicate that, under ideal or low-noise conditions and particularly for small-scale problems, ans\"atze with high Hamiltonian expressibility yield better performance for problems with non-diagonal Hamiltonians and superposition-state solutions. Conversely, circuits with low expressibility are more effective for problems whose solutions are basis states, including those defined by diagonal Hamiltonians. Under noisy conditions, low-expressibility circuits remain preferable for basis-state problems, while intermediate expressibility yields better results for some problems involving superposition-state solutions.

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

75 extracted references · 18 canonical work pages

  1. [1]

    J Supercomput 81(5):687

    AbuGhanem M (2025) Ibm quantum computers: evolution, performance, and future directions. J Supercomput 81(5):687. doi:10.1007/s11227-025-07047-7

    doi:10.1007/s11227-025-07047-7 2025

  2. [2]

    Rev Mod Phys 90(1):015002

    Albash T, Lidar DA (2018) Adiabatic quantum computation. Rev Mod Phys 90(1):015002. doi:10.1103/revmodphys.90.015002

    doi:10.1103/revmodphys.90.015002 2018

  3. [3]

    Quantum 5:558

    Arrasmith A, Cerezo M, Czarnik P, et al (2021) Effect of barren plateaus on gradient-free optimization. Quantum 5:558. doi:10.22331/q-2021-10-05-558

    doi:10.22331/q-2021-10-05-558 2021

  4. [4]

    Quantum Sci Technol 7(4):045015

    Arrasmith A, Holmes Z, Cerezo M, et al (2022) Equivalence of quantum barren plateaus to cost concentration and narrow gorges. Quantum Sci Technol 7(4):045015. doi:10.1088/2058-9565/ac7d06

    doi:10.1088/2058-9565/ac7d06 2022

  5. [5]

    https://arxiv.org/abs/2103.06712 arXiv:2103.06712

    Bilkis M, Cerezo M, Verdon G, et al (2023) A semi-agnostic ansatz with variable structure for quantum machine learning. https://arxiv.org/abs/2103.06712 arXiv:2103.06712

    arXiv 2023

  6. [6]

    Phys Rep 1068:1--66

    Blekos K, Brand D, Ceschini A, et al (2024) A review on quantum approximate optimization algorithm and its variants. Phys Rep 1068:1--66. doi:https://doi.org/10.1016/j.physrep.2024.03.002

    doi:10.1016/j.physrep.2024.03.002 2024

  7. [7]

    Phys Lett A 294(4):258--260

    Bowdrey M, Oi D, Short A, et al (2002) Fidelity of single qubit maps. Phys Lett A 294(4):258--260. doi:10.1016/S0375-9601(01)00745-2

    work page doi:10.1016/s0375-9601(01)00745-2 2002

  8. [8]

    Chem Rev 119(19):10856--10915

    Cao Y, Romero J, Olson JP, et al (2019) Quantum chemistry in the age of quantum computing. Chem Rev 119(19):10856--10915. doi:10.1021/acs.chemrev.8b00803

    doi:10.1021/acs.chemrev.8b00803 2019

Show all 75 references
  1. [9]

    Quantum Sci Technol 6(3):035006

    Cerezo M, Coles PJ (2021) Higher order derivatives of quantum neural networks with barren plateaus. Quantum Sci Technol 6(3):035006. doi:10.1088/2058-9565/abf51a

    2021 doi

  2. [10]

    Nat Rev Phys 3(9)

    Cerezo M, Arrasmith A, Babbush R, et al (2021) Variational quantum algorithms. Nat Rev Phys 3(9). doi:10.1038/s42254-021-00348-9

    2021 doi

  3. [11]

    https://arxiv.org/abs/2007.04424 arXiv:2007.04424

    Chivilikhin D, Samarin A, Ulyantsev V, et al (2020) Mog-vqe: Multiobjective genetic variational quantum eigensolver. https://arxiv.org/abs/2007.04424 arXiv:2007.04424

    2020 arXiv

  4. [12]

    New J Phys 20(11):113022

    Cincio L, Suba s Y, Sornborger AT, et al (2018) Learning the quantum algorithm for state overlap. New J Phys 20(11):113022. doi:10.1088/1367-2630/aae94a

    2018 doi

  5. [13]

    Front Chem 8

    Claudino D, Wright J, McCaskey AJ, et al (2020) Benchmarking adaptive variational quantum eigensolvers. Front Chem 8. doi:10.3389/fchem.2020.606863

    2020

  6. [14]

    Commun Math Phys 264(3):773--795

    Collins B, Śniady P (2006) Integration with respect to the haar measure on unitary, orthogonal and symplectic group. Commun Math Phys 264(3):773--795. doi:10.1007/s00220-006-1554-3

    2006 doi

  7. [15]

    Wiley-Interscience, Hoboken, NJ

    Cover TM, Thomas JA (2006) Elements of Information Theory, 2nd edn. Wiley-Interscience, Hoboken, NJ

    2006

  8. [16]

    Phys Rev Res 2:033125

    Du Y, Hsieh MH, Liu T, et al (2020) Expressive power of parametrized quantum circuits. Phys Rev Res 2:033125. doi:10.1103/PhysRevResearch.2.033125

    2020 doi

  9. [17]

    npj Quantum Inf 8(1)

    Du Y, Huang T, You S, et al (2022 a ) Quantum circuit architecture search for variational quantum algorithms. npj Quantum Inf 8(1). doi:10.1038/s41534-022-00570-y

    2022 doi

  10. [18]

    Phys Rev Lett 128(8)

    Du Y, Tu Z, Yuan X, et al (2022 b ) Efficient measure for the expressivity of variational quantum algorithms. Phys Rev Lett 128(8). doi:10.1103/physrevlett.128.080506

    2022 doi

  11. [19]

    Nature 622(7982):268--272

    Evered SJ, Bluvstein D, Kalinowski M, et al (2023) High-fidelity parallel entangling gates on a neutral-atom quantum computer. Nature 622(7982):268--272. doi:10.1038/s41586-023-06481-y

    2023 doi

  12. [20]

    https://arxiv.org/abs/1411.4028 arXiv:1411.4028

    Farhi E, Goldstone J, Gutmann S (2014) A quantum approximate optimization algorithm. https://arxiv.org/abs/1411.4028 arXiv:1411.4028

    2014 arXiv

  13. [21]

    A.and Peng, Govind N, Alexeev Y (2022) Vqe method: a short survey and recent developments

    Fedorov BD. A.and Peng, Govind N, Alexeev Y (2022) Vqe method: a short survey and recent developments. Mater Theory 6:2. doi:10.1186/s41313-021-00032-6

    2022 doi

  14. [22]

    Commun Phys 6(1):192

    Feniou C, Hassan M, Traoré D, et al (2023) Overlap-adapt-vqe: practical quantum chemistry on quantum computers via overlap-guided compact ansätze. Commun Phys 6(1):192. doi:10.1038/s42005-023-01312-y

    2023 doi

  15. [23]

    Preprint at https://arxiv.org/abs/2409.05475, https://arxiv.org/abs/2409.05475 arXiv:2409.05475

    Foderà S, Turati G, Nembrini R, et al (2024) Reinforcement learning for variational quantum circuits design. Preprint at https://arxiv.org/abs/2409.05475, https://arxiv.org/abs/2409.05475 arXiv:2409.05475

    2024

  16. [24]

    Phys Rev A 104(2)

    Fontana E, Fitzpatrick N, Ramo DM, et al (2021) Evaluating the noise resilience of variational quantum algorithms. Phys Rev A 104(2). doi:10.1103/physreva.104.022403

    2021 doi

  17. [25]

    Quantum 5:422

    Funcke L, Hartung T, Jansen K, et al (2021) Dimensional expressivity analysis of parametric quantum circuits. Quantum 5:422. doi:10.22331/q-2021-03-29-422

    2021 doi

  18. [26]

    Phys Rev Appl 10:024052

    Geller MR (2018) Sampling and scrambling on a chain of superconducting qubits. Phys Rev Appl 10:024052. doi:10.1103/PhysRevApplied.10.024052

    2018 doi

  19. [27]

    ://arxiv.org/abs/2311.01330, https://arxiv.org/abs/2311.01330 arXiv:2311.01330

    Ghosh T, Mandal A, Banerjee S, et al (2023) Lower bound of the expressibility of ansatzes for variational quantum algorithms. ://arxiv.org/abs/2311.01330, https://arxiv.org/abs/2311.01330 arXiv:2311.01330

    2023 arXiv

  20. [28]

    Phys Rev Res 4:043056

    Giordano S, Martin-Delgado MA (2022) Reinforcement-learning generation of four-qubit entangled states. Phys Rev Res 4:043056. doi:10.1103/PhysRevResearch.4.043056

    2022 doi

  21. [29]

    Phys Rev Lett 122:127202

    Gopalakrishnan S, Vasseur R (2019) Kinetic theory of spin diffusion and superdiffusion in xxz spin chains. Phys Rev Lett 122:127202. doi:10.1103/PhysRevLett.122.127202

    2019 doi

  22. [30]

    Nat Commun 10(1)

    Grimsley HR, Economou SE, Barnes E, et al (2019) An adaptive variational algorithm for exact molecular simulations on a quantum computer. Nat Commun 10(1). doi:10.1038/s41467-019-10988-2

    2019 doi

  23. [31]

    Phys Rev Lett 116(23)

    Heras UL, Alvarez-Rodriguez U, Solano E, et al (2016) Genetic algorithms for digital quantum simulations. Phys Rev Lett 116(23). doi:10.1103/physrevlett.116.230504

    2016 doi

  24. [32]

    PRX Quantum 3:010313

    Holmes Z, Sharma K, Cerezo M, et al (2022) Connecting ansatz expressibility to gradient magnitudes and barren plateaus. PRX Quantum 3:010313. doi:10.1103/prxquantum.3.010313

    2022 doi

  25. [33]

    ://arxiv.org/abs/1711.02249, https://arxiv.org/abs/1711.02249 arXiv:1711.02249

    Johnson PD, Romero J, Olson J, et al (2017) Qvector: an algorithm for device-tailored quantum error correction. ://arxiv.org/abs/1711.02249, https://arxiv.org/abs/1711.02249 arXiv:1711.02249

    2017 arXiv

  26. [34]

    Nature 549(7671):242--246

    Kandala A, Mezzacapo A, Temme K, et al (2017) Hardware-efficient variational quantum eigensolver for small molecules and quantum magnets. Nature 549(7671):242--246. doi:10.1038/nature23879

    2017 doi

  27. [35]

    Biometrika 30(1-2):81--93

    Kendall MG (1938) A new measure of rank correlation. Biometrika 30(1-2):81--93. doi:10.1093/biomet/30.1-2.81

    1938 doi

  28. [36]

    https://arxiv.org/abs/2104.07715 arXiv:2104.07715

    Kuo EJ, Fang YLL, Chen SYC (2021) Quantum architecture search via deep reinforcement learning. https://arxiv.org/abs/2104.07715 arXiv:2104.07715

    2021 arXiv

  29. [37]

    Quantum 6:824

    Larocca M, Czarnik P, Sharma K, et al (2022) Diagnosing B arren P lateaus with T ools from Q uantum O ptimal C ontrol. Quantum 6:824. doi:10.22331/q-2022-09-29-824

    2022 doi

  30. [38]

    ://arxiv.org/abs/2407.07694, https://arxiv.org/abs/2407.07694 arXiv:2407.07694

    Löschnauer CM, Toba JM, Hughes AC, et al (2024) Scalable, high-fidelity all-electronic control of trapped-ion qubits. ://arxiv.org/abs/2407.07694, https://arxiv.org/abs/2407.07694 arXiv:2407.07694

    2024

  31. [39]

    New J Phys 18(2):023023

    McClean JR, Romero J, Babbush R, et al (2016) The theory of variational hybrid quantum-classical algorithms. New J Phys 18(2):023023. doi:10.1088/1367-2630/18/2/023023

    2016 doi

  32. [40]

    Nat Commun 9(1)

    McClean JR, Boixo S, Smelyanskiy VN, et al (2018) Barren plateaus in quantum neural network training landscapes. Nat Commun 9(1). doi:10.1038/s41467-018-07090-4

    2018 doi

  33. [41]

    Cambridge Tracts in Mathematics, Cambridge University Press, Cambridge

    Meckes ES (2019) The Random Matrix Theory of the Classical Compact Groups. Cambridge Tracts in Mathematics, Cambridge University Press, Cambridge

    2019

  34. [42]

    Quantum 8:1340

    Mele AA (2024) Introduction to haar measure tools in quantum information: A beginner's tutorial. Quantum 8:1340. doi:10.22331/q-2024-05-08-1340

    2024 doi

  35. [43]

    PRX Quantum 4:010328

    Meyer JJ, Mularski M, Gil-Fuster E, et al (2023) Exploiting symmetry in variational quantum machine learning. PRX Quantum 4:010328. doi:10.1103/PRXQuantum.4.010328

    2023 doi

  36. [44]

    PRX Quantum 2:040316

    Ortiz Marrero C, Kieferov\'a M, Wiebe N (2021) Entanglement-induced barren plateaus. PRX Quantum 2:040316. doi:10.1103/PRXQuantum.2.040316

    2021 doi

  37. [45]

    Quantum 5:391

    Ostaszewski M, Grant E, Benedetti M (2021 a ) Structure optimization for parameterized quantum circuits. Quantum 5:391. doi:10.22331/q-2021-01-28-391

    2021 doi

  38. [46]

    In: Advances in Neural Information Processing Systems, pp 18182--18194

    Ostaszewski M, Trenkwalder LM, Masarczyk W, et al (2021 b ) Reinforcement learning for optimization of variational quantum circuit architectures. In: Advances in Neural Information Processing Systems, pp 18182--18194

    2021

  39. [47]

    Pearson K (1896) Mathematical contributions to the theory of evolution. iii. regression, heredity, and panmixia. Philosophical Transactions of the Royal Society of London Series A, Containing Papers of a Mathematical or Physical Character 187:253--318. doi:10.1098/rsta.1896.0007

  40. [48]

    Nat Commun 5(1)

    Peruzzo A, McClean J, Shadbolt P, et al (2014) A variational eigenvalue solver on a photonic quantum processor. Nat Commun 5(1). doi:10.1038/ncomms5213

    2014 doi

  41. [49]

    Ann Phys 57:79--90

    Pfeuty P (1970) The one-dimensional king model with a transverse field. Ann Phys 57:79--90. doi:https://doi.org/10.1016/0003-4916(70)90270-8

    1970 doi

  42. [50]

    Quantum Mach Intell 3(2):25

    Pirhooshyaran M, Terlaky T (2021) Quantum circuit design search. Quantum Mach Intell 3(2):25. doi:10.1007/s42484-021-00051-z

    2021 doi

  43. [51]

    In: Advances in Optimization and Numerical Analysis

    Powell MJD (1994) A direct search optimization method that models the objective and constraint functions by linear interpolation. In: Advances in Optimization and Numerical Analysis. Springer Netherlands, Heidelberg, p 51--67

    1994

  44. [52]

    Quantum 2:79

    Preskill J (2018) Quantum computing in the nisq era and beyond. Quantum 2:79. doi:10.22331/q-2018-08-06-79

    2018 doi

  45. [53]

    J Phys Conf Ser 2634(1):012043

    Qin J (2023) Review of ansatz designing techniques for variational quantum algorithms. J Phys Conf Ser 2634(1):012043. doi:10.1088/1742-6596/2634/1/012043

    2023 doi

  46. [54]

    https://arxiv.org/abs/1910.09694 arXiv:1910.09694

    Rattew AG, Hu S, Pistoia M, et al (2020) A domain-agnostic, noise-resistant, hardware-efficient evolutionary variational quantum eigensolver. https://arxiv.org/abs/1910.09694 arXiv:1910.09694

    2020 arXiv

  47. [55]

    J High Energy Phys 2017(4):121

    Roberts DA, Yoshida B (2017) Chaos and complexity by design. J High Energy Phys 2017(4):121. doi:10.1007/JHEP04(2017)121

    2017 doi

  48. [56]

    Quantum Sci Technol 2(4):045001

    Romero J, Olson JP, Aspuru-Guzik A (2017) Quantum autoencoders for efficient compression of quantum data. Quantum Sci Technol 2(4):045001. doi:10.1088/2058-9565/aa8072

    2017 doi

  49. [57]

    Phys Rev A 101(3)

    Schuld M, Bocharov A, Svore KM, et al (2020) Circuit-centric quantum classifiers. Phys Rev A 101(3). doi:10.1103/physreva.101.032308

    2020 doi

  50. [58]

    Quantum Sci Technol 7(2):023002

    Shi XF (2022) Quantum logic and entanglement by neutral rydberg atoms: methods and fidelity. Quantum Sci Technol 7(2):023002. doi:10.1088/2058-9565/ac18b8

    2022 doi

  51. [59]

    Adv Quantum Technol 2(12)

    Sim S, Johnson PD, Aspuru‐Guzik A (2019) Expressibility and entangling capability of parameterized quantum circuits for hybrid quantum‐classical algorithms. Adv Quantum Technol 2(12). doi:10.1002/qute.201900070

    2019 doi

  52. [60]

    American Mathematical Society, Providence

    Simon B (1996) Representations of Finite and Compact Groups, vol 10. American Mathematical Society, Providence

    1996

  53. [61]

    ://arxiv.org/abs/quant-ph/0602174, https://arxiv.org/abs/quant-ph/0602174 arXiv:quant-ph/0602174

    Sousa PBM, Ramos RV (2006) Universal quantum circuit for n-qubit quantum gate: A programmable quantum gate. ://arxiv.org/abs/quant-ph/0602174, https://arxiv.org/abs/quant-ph/0602174 arXiv:quant-ph/0602174

    2006 arXiv

  54. [62]

    Am J Psychol 15(1):72--101

    Spearman C (1904) The proof and measurement of association between two things. Am J Psychol 15(1):72--101. doi:10.2307/1412159

    1904 doi

  55. [63]

    PRX Quantum 2(2)

    Tang HL, Shkolnikov V, Barron GS, et al (2021) Qubit- ADAPT - VQE : An adaptive algorithm for constructing hardware-efficient ansätze on a quantum processor. PRX Quantum 2(2). doi:10.1103/prxquantum.2.020310

    2021 doi

  56. [64]

    Phys Rep 986:1--128

    Tilly J, Chen H, Cao S, et al (2022) The variational quantum eigensolver: A review of methods and best practices. Phys Rep 986:1--128. doi:https://doi.org/10.1016/j.physrep.2022.08.003

    2022 doi

  57. [65]

    In: 2023 IEEE International Conference on Quantum Computing and Engineering (QCE), pp 407--413, doi:10.1109/QCE57702.2023.00053

    Turati G, Ferrari Dacrema M, Cremonesi P (2023) Benchmarking adaptative variational quantum algorithms on QUBO instances. In: 2023 IEEE International Conference on Quantum Computing and Engineering (QCE), pp 407--413, doi:10.1109/QCE57702.2023.00053

    2023

  58. [66]

    Quantum Sci Technol 6(2):025008

    Volkoff T, Coles PJ (2021) Large gradients via correlation in random parameterized quantum circuits. Quantum Sci Technol 6(2):025008. doi:10.1088/2058-9565/abd891

    2021 doi

  59. [67]

    ://arxiv.org/abs/1806.08321, https://arxiv.org/abs/1806.08321 arXiv:1806.08321

    Wilson CM, Otterbach JS, Tezak N, et al (2019) Quantum kitchen sinks: An algorithm for machine learning on near-term quantum computers. ://arxiv.org/abs/1806.08321, https://arxiv.org/abs/1806.08321 arXiv:1806.08321

    2019 arXiv

  60. [68]

    ://arxiv.org/abs/2111.13730, https://arxiv.org/abs/2111.13730 arXiv:2111.13730

    Wu A, Li G, Wang Y, et al (2021) Towards efficient ansatz architecture for variational quantum algorithms. ://arxiv.org/abs/2111.13730, https://arxiv.org/abs/2111.13730 arXiv:2111.13730

    2021 arXiv

  61. [69]

    IEEE Trans Quantum Eng 2:1--7

    Wurtz J, Love PJ (2021) Classically optimal variational quantum algorithms. IEEE Trans Quantum Eng 2:1--7. doi:10.1109/TQE.2021.3122568

    2021

  62. [70]

    Commun Phys 4(1)

    Yordanov YS, Armaos V, Barnes CHW, et al (2021) Qubit-excitation-based adaptive variational quantum eigensolver. Commun Phys 4(1). doi:10.1038/s42005-021-00730-0

    2021 doi

  63. [71]

    https://arxiv.org/abs/2005.10258 arXiv:2005.10258

    Zhu L, Tang HL, Barron GS, et al (2022) An adaptive quantum approximate optimization algorithm for solving combinatorial problems on a quantum computer. https://arxiv.org/abs/2005.10258 arXiv:2005.10258

    2022 arXiv

  64. [72]

    Sci Rep 13(1):5157

    Zhu X, Hou X (2023) Quantum architecture search via truly proximal policy optimization. Sci Rep 13(1):5157. doi:10.1038/s41598-023-32349-2

    2023 doi

  65. [73]

    sn-basic.bst

    FUNCTION identify.basic.version "sn-basic.bst" " [2024/07/19 v1.1 bibliography style]" * top ENTRY address archive author booktitle chapter doi edition editor eid eprint howpublished institution journal key keywords month note number organization pages publisher school series ...

    2024

  66. [74]

    write newline

    " write newline "" before.all 'output.state := FUNCTION add.period duplicate empty 'skip "." * add.blank if FUNCTION if.digit duplicate "0" = swap duplicate "1" = swap duplicate "2" = swap duplicate "3" = swap duplicate "4" = swap duplicate "5" = swap duplicate "6" = swap dupl...

  67. [75]

    write newline

    " write newline "" before.all 'output.state := FUNCTION string.to.integer 't := t text.length 'k := #1 'char.num := t char.num #1 substring 's := s is.num s "." = or char.num k = not and char.num #1 + 'char.num := while char.num #1 - 'char.num := t #1 char.num substring FUNCTI...

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.