pith. sign in

arxiv: 2507.18844 · v2 · submitted 2025-07-24 · 🪐 quant-ph · cs.ET

Exploring Entanglement and Parameter Sensitivity in QAOA through Quantum Fisher Information

Brian Garc\'ia Sarmina , Jorge Saavedra Benavides , Guo-Hua Sun , Shi-Hai Dong This is my paper

Pith reviewed 2026-05-19 02:10 UTC · model grok-4.3

classification 🪐 quant-ph cs.ET
keywords Quantum Fisher InformationQAOAMax-CutVariational Quantum AlgorithmsEntanglementParameter SensitivityOptimization Heuristics
0
0 comments X

The pith

Quantum Fisher Information supplies mutation rates that lift QAOA performance above random baselines on small Max-Cut graphs.

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

The paper measures how QAOA states respond to small changes in their parameters by computing the Quantum Fisher Information matrix for circuits of depth up to nine on graphs with four to ten qubits. It finds that complete graphs produce larger sensitivity eigenvalues than cyclic graphs, while added entanglement layers move sensitivity from single parameters into cross-parameter correlations. From these patterns the authors build a simple proof-of-concept rule that reads the normalized diagonal QFI entries to choose which parameters to mutate and by how much. When tested on seven- and ten-qubit instances this rule returns higher average cut values and smaller run-to-run spread than equal-probability or random-restart strategies across one hundred trials. The result positions QFI as a lightweight, circuit-aware guide for tuning variational quantum optimizers.

Core claim

Complete graphs yield larger QFI eigenvalues than cyclic graphs; entanglement redistributes weight from diagonal to off-diagonal matrix entries; and a mutation heuristic that sets probabilities and step sizes from the normalized diagonal QFI produces higher mean energies and lower variance than equal-probability or random-restart baselines on seven- and ten-qubit Max-Cut problems.

What carries the argument

The QFI-Informed Mutation (QIm) heuristic, which converts the normalized diagonal elements of the Quantum Fisher Information matrix into per-parameter mutation probabilities and step sizes.

If this is right

  • Entangling stages increase the fraction of QFI that resides in off-diagonal covariances, raising cross-parameter correlations.
  • None of the tested circuit families reaches the Heisenberg limit of 4N squared, yet several exceed the linear bound of 4N.
  • The mutation rule functions as a problem-aware preconditioner that lowers outcome variance without requiring additional circuit executions beyond the QFI calculation.
  • QFI analysis can be repeated for other variational quantum algorithms to obtain similar parameter-sensitivity diagnostics.

Where Pith is reading between the lines

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

  • The same diagonal-QFI rule might be used to choose initial parameter values rather than only mutation steps during search.
  • Comparing QFI spectra across graph families could indicate which problem structures benefit most from early entanglement stages.
  • Extending the heuristic from discrete mutations to continuous gradient steps would test whether the sensitivity information remains useful in gradient-based VQA training.

Load-bearing premise

The normalized diagonal QFI values obtained on the studied small instances and circuit families supply mutation probabilities and step sizes that improve optimization performance.

What would settle it

Running the same QIm heuristic on QAOA instances larger than ten qubits or on different combinatorial problems and checking whether it still outperforms the equal-probability and random-restart baselines.

Figures

Figures reproduced from arXiv: 2507.18844 by Brian Garc\'ia Sarmina, Guo-Hua Sun, Jorge Saavedra Benavides, Shi-Hai Dong.

Figure 1
Figure 1. Figure 1: FIG. 1: One-layer QAOA circuits on 4 qubits with RX-only and R [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2: One-layer QAOA circuits on 4 qubits with different en [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3: Cyclic and complete 7-node Max-Cut instances used as [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4: RX-only mixing operators for cyclic max-cut problem [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5: RX-only mixing operators for complete max-cut probl [PITH_FULL_IMAGE:figures/full_fig_p007_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6: RX-RY mixing operators for cyclic max-cut problems w [PITH_FULL_IMAGE:figures/full_fig_p007_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7: RX-RY mixing operators for complete max-cut problem [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8: Covariance fraction for the cyclic and complete max- [PITH_FULL_IMAGE:figures/full_fig_p008_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: FIG. 9: Covariance fraction for the cyclic and complete max- [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: FIG. 10: Maximum and minimum eigenvalues for cyclic vs. comp [PITH_FULL_IMAGE:figures/full_fig_p009_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: FIG. 11: Covariance fraction for cyclic vs. complete entang [PITH_FULL_IMAGE:figures/full_fig_p010_11.png] view at source ↗
Figure 12
Figure 12. Figure 12: FIG. 12: Maximum and minimum eigenvalues for the cyclic and c [PITH_FULL_IMAGE:figures/full_fig_p010_12.png] view at source ↗
Figure 13
Figure 13. Figure 13: FIG. 13: Covariance fraction for the cyclic and complete ent [PITH_FULL_IMAGE:figures/full_fig_p011_13.png] view at source ↗
Figure 14
Figure 14. Figure 14: FIG. 14: Results of 100 simulations for QIm, nonQIm, and RR on [PITH_FULL_IMAGE:figures/full_fig_p012_14.png] view at source ↗
Figure 15
Figure 15. Figure 15: FIG. 15: QFI matrices for the 7-node max-cut complete configu [PITH_FULL_IMAGE:figures/full_fig_p016_15.png] view at source ↗
Figure 16
Figure 16. Figure 16: FIG. 16: QFI matrix for the 10-node max-cut cyclic configurat [PITH_FULL_IMAGE:figures/full_fig_p016_16.png] view at source ↗
read the original abstract

Quantum Fisher Information (QFI) can be used to quantify how sensitive a quantum state reacts to changes in its variational parameters, making it a natural diagnostic for algorithms such as the Quantum Approximate Optimization Algorithm (QAOA). We perform a systematic QFI analysis of QAOA for Max-Cut on cyclic and complete graphs with $N = 4 - 10$ qubits. Two mixer families are studied, RX-only and hybrid RX-RY, with depths $p = 2, 4, 6$ and $p = 3, 6, 9$, respectively, and with up to three entanglement stages implemented through cyclic- or complete-entangling patterns. Complete graphs consistently yield larger QFI eigenvalues than cyclic graphs; none of the settings reaches the Heisenberg limit ($4N^2$), but several exceed the linear bound ($4N$). Introducing entanglement primarily redistributes QFI from diagonal to off-diagonal entries: non-entangled circuits maximize per-parameter (diagonal) sensitivity, whereas entangling layers increase the covariance fraction and thus cross-parameter correlations, with diminishing returns beyond the first stage. Leveraging these observations, we propose, as a proof of concept, a QFI-Informed Mutation (QIm) heuristic that sets mutation probabilities and step sizes from the normalized diagonal QFI. On 7- and 10-qubit instances, QIm attains higher mean energies and lower variance than equal-probability and random-restart baselines over 100 runs, underscoring QFI as a lightweight, problem-aware preconditioner for QAOA and other variational quantum algorithms.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

2 major / 2 minor

Summary. The manuscript conducts a systematic numerical study of the Quantum Fisher Information (QFI) matrix for QAOA applied to Max-Cut on cyclic and complete graphs with N=4–10 qubits. It compares RX-only and hybrid RX-RY mixer families at depths p=2,4,6 and p=3,6,9 respectively, with up to three entanglement stages using cyclic or complete entangling patterns. Findings include consistently larger QFI eigenvalues on complete graphs, several settings exceeding the linear bound 4N but none reaching the Heisenberg limit 4N², and entanglement primarily shifting QFI weight from diagonal to off-diagonal entries with diminishing returns after the first stage. From these diagnostics the authors introduce a proof-of-concept QFI-Informed Mutation (QIm) heuristic that derives mutation probabilities and step sizes from the normalized diagonal QFI; on 7- and 10-qubit instances this heuristic reports higher mean energies and lower variance than equal-probability and random-restart baselines across 100 runs.

Significance. If the reported performance gains prove robust, the work would usefully position QFI as a lightweight, problem-aware diagnostic and preconditioner for variational quantum algorithms. The systematic exploration of entanglement effects on QFI redistribution for small but representative Max-Cut instances supplies concrete, falsifiable observations that can guide circuit design. The heuristic itself is presented as a proof-of-concept rather than a fully optimized method, so its broader impact hinges on statistical validation and scaling behavior.

major comments (2)
  1. [QIm heuristic performance evaluation] In the section presenting the QIm heuristic results, the abstract and associated text state that QIm attains higher mean energies and lower variance than the two baselines over 100 runs on 7- and 10-qubit instances, yet no standard errors, confidence intervals, or hypothesis-test statistics (e.g., paired t-test or Wilcoxon rank-sum p-values) are reported. Because the underlying optimization is stochastic, these omissions leave open the possibility that the observed differences lie within run-to-run variation; this is load-bearing for the central claim that the normalized diagonal QFI supplies an effective preconditioner.
  2. [QFI eigenvalue analysis] The claim that “none of the settings reaches the Heisenberg limit (4N²)” and that “several exceed the linear bound (4N)” is presented without an explicit table or figure listing the largest eigenvalues for each (N, p, mixer, entanglement) combination. Adding such a summary table would allow direct verification that the reported exceedances are not artifacts of numerical precision or particular graph realizations.
minor comments (2)
  1. [Abstract / Methods] The abstract refers to “up to three entanglement stages implemented through cyclic- or complete-entangling patterns” but does not specify the exact two-qubit gate sequence or the placement of these stages within the QAOA circuit; a short methods paragraph or supplementary circuit diagram would improve reproducibility.
  2. [Heuristic definition] Notation for the normalized diagonal QFI used to set mutation probabilities should be introduced once with an explicit formula (e.g., p_i = F_ii / sum_j F_jj) rather than described only qualitatively, to avoid ambiguity when readers attempt to re-implement the heuristic.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their thorough review and valuable suggestions. We address each of the major comments below and indicate the revisions we will make to the manuscript.

read point-by-point responses

  1. Referee: [QIm heuristic performance evaluation] In the section presenting the QIm heuristic results, the abstract and associated text state that QIm attains higher mean energies and lower variance than the two baselines over 100 runs on 7- and 10-qubit instances, yet no standard errors, confidence intervals, or hypothesis-test statistics (e.g., paired t-test or Wilcoxon rank-sum p-values) are reported. Because the underlying optimization is stochastic, these omissions leave open the possibility that the observed differences lie within run-to-run variation; this is load-bearing for the central claim that the normalized diagonal QFI supplies an effective preconditioner.

    Authors: We agree with the referee that statistical measures are necessary to substantiate the performance claims of the QIm heuristic given the stochastic nature of the optimization. In the revised version, we will report standard errors for the reported means and variances. Additionally, we will include p-values from paired t-tests or non-parametric alternatives to evaluate whether the improvements are statistically significant. These additions will be incorporated into the results section and referenced in the abstract. We have performed the necessary post-processing on our existing run data to generate these statistics. revision: yes

  2. Referee: [QFI eigenvalue analysis] The claim that “none of the settings reaches the Heisenberg limit (4N²)” and that “several exceed the linear bound (4N)” is presented without an explicit table or figure listing the largest eigenvalues for each (N, p, mixer, entanglement) combination. Adding such a summary table would allow direct verification that the reported exceedances are not artifacts of numerical precision or particular graph realizations.

    Authors: We appreciate this suggestion for improving the transparency of our QFI eigenvalue analysis. We will add a summary table to the manuscript that explicitly lists the largest QFI eigenvalue for every combination of system size N, circuit depth p, mixer family, and entanglement pattern. The table will also flag which values exceed the linear bound of 4N. This will enable readers to verify our statements directly from the numerical results without ambiguity. revision: yes

Circularity Check

0 steps flagged

No circularity: QFI diagnostic informs heuristic construction but performance claims rest on independent empirical runs

full rationale

The paper first computes the quantum Fisher information matrix directly from the QAOA circuit unitaries for the chosen depths, mixers, and graph topologies. It then observes patterns (e.g., redistribution of QFI from diagonal to off-diagonal entries upon adding entanglement) and uses the normalized diagonal QFI values to define the mutation probabilities and step sizes of the QIm heuristic. The central performance claim—that QIm yields higher mean energies and lower variance than baselines—is established by running the heuristic and the control methods on 7- and 10-qubit instances over 100 independent trials and comparing the resulting energy distributions. This empirical comparison does not reduce to any equation or fitted parameter inside the QFI analysis; the heuristic is an externally constructed procedure whose effectiveness is tested against separate stochastic baselines. No self-citations, uniqueness theorems, or ansatzes are invoked in a load-bearing manner, and the derivation chain remains self-contained against the reported numerical benchmarks.

Axiom & Free-Parameter Ledger

3 free parameters · 2 axioms · 0 invented entities

The central claim rests on standard quantum information concepts and numerical choices for small-system QAOA. No new physical entities are postulated. Free parameters are the discrete choices of circuit depths, entanglement stage counts, and graph connectivities selected for the study.

free parameters (3)
  • Circuit depth p
    Selected values 2,4,6 for RX-only and 3,6,9 for RX-RY mixers
  • Number of entanglement stages
    Up to three stages with cyclic or complete patterns
  • Graph family
    Cyclic versus complete graphs for Max-Cut instances
axioms (2)
  • domain assumption QFI quantifies parameter sensitivity of variational quantum states
    Invoked as natural diagnostic for QAOA parameter landscape
  • standard math Max-Cut on graphs is a standard QAOA benchmark problem
    Used to define the objective for the studied instances

pith-pipeline@v0.9.0 · 5830 in / 1615 out tokens · 96679 ms · 2026-05-19T02:10:04.813944+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

What do these tags mean?
matches
The paper's claim is directly supported by a theorem in the formal canon.
supports
The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends
The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses
The paper appears to rely on the theorem as machinery.
contradicts
The paper's claim conflicts with a theorem or certificate in the canon.
unclear
Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

45 extracted references · 45 canonical work pages · 2 internal anchors

  1. [1]

    Random-parameter sampling: For each setting, 100 random parameter draws in [0, 2π ) are used to probe the nonuniform sensitivity landscape; QFI matrices are averaged to suppre ss sampling noise

    work page

  2. [2]

    Entanglement-pattern comparison: We contrast two CNOT-based patterns in the mixer, cyclic (nearest-neighbor ring) and complete (all-to-all), to isolate the effect of entanglement connec tivity on QFI

    work page

  3. [3]

    This study is carried out on the 7-node complete graph

    Entanglement-stage count: Holding the overall depth fixed (e.g., 3L), we vary the number of entanglement stages (1, 2, or 3) to assess how repeated entangling layers redistribu te QFI between diagonal and off-diagonal components. This study is carried out on the 7-node complete graph. IV . RESULTS AND DISCUSSION We report Quantum Fisher Information (QFI) f...

    work page

  4. [4]

    55 → 0. 51 → 0. 58, cyclic: 0. 49 → 0. 42 → 0. 39). Thus, while the first entangling layer delivers the larges t QFI gain and cross-talk increase, additional stages offer diminish ing and sometimes non-monotonic returns. The key findings of the QFI results of the different QAOA and entanglement models are su mmarized in Table I. 10 2 3 4 5 6 Parameters p 0....

    work page

  5. [5]

    A Quantum Approximate Optimization Algorithm

    E. Farhi, J. Goldstone, and S. Gutmann, “A quantum approx imate optimization algorithm,” arXiv preprint arXiv:1411.4028, 2014

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  6. [6]

    Quantum approximate optimization of non-planar graph problems on a planar superconducting pr ocessor,

    M. P . e. a. Harrigan, “Quantum approximate optimization of non-planar graph problems on a planar superconducting pr ocessor,” Nature Physics, vol. 17, pp. 332–336, 2021

    work page 2021

  7. [7]

    Q uantum approximate optimization algorithm: Performance, mechanism, and implementation on near-term devices,

    L. Zhou, S. T. Wang, S. Choi, H. Pichler, and M. D. Lukin, “Q uantum approximate optimization algorithm: Performance, mechanism, and implementation on near-term devices,” Physical Review X, vol. 10, no. 2, p. 021067, 2020

    work page 2020

  8. [8]

    Quantum approximate optimization algorithm for max-cut: A fermion ic view,

    Z. Wang, S. Hadfield, Z. Jiang, and E. G. Rieffel, “Quantum approximate optimization algorithm for max-cut: A fermion ic view,” Physical Review A, vol. 97, no. 2, p. 022304, 2018

    work page 2018

  9. [9]

    A tutorial on quantum approximate opt imization algorithm (qaoa): Fundamentals and application s,

    J. Choi and J. Kim, “A tutorial on quantum approximate opt imization algorithm (qaoa): Fundamentals and application s,” in 2019 international conference on information and communicatio n technology convergence (ICTC). IEEE, 2019, pp. 138–142

    work page 2019

  10. [10]

    Barren plateaus in quantum neural network training landscapes,

    J. R. McClean, S. Boixo, V . N. Smelyanskiy, R. Babbush, and H. Neven, “Barren plateaus in quantum neural network training landscapes,” Nature Communications, vol. 9, no. 1, p. 4812, 2018

    work page 2018

  11. [11]

    Noise-induced barren plateaus in varia tional quantum algorithms,

    S. Wang, E. Fontana, M. Cerezo, K. Sharma, A. Sone, L. Cinc io, and P . J. Coles, “Noise-induced barren plateaus in varia tional quantum algorithms,” Nature Communications, vol. 12, no. 1, p. 6961, 2021

    work page 2021

  12. [12]

    Comparison of QAOA with Quantum and Simulated Annealing

    M. Streif and M. Leib, “Comparison of qaoa with quantum an d simulated annealing,” arXiv preprint arXiv:1901.01903, 2019

    work page internal anchor Pith review Pith/arXiv arXiv 1901

  13. [13]

    Qaoa-in-qaoa: solvin g large-scale maxcut problems on small quantum machines,

    Z. Zhou, Y . Du, X. Tian, and D. Tao, “Qaoa-in-qaoa: solvin g large-scale maxcut problems on small quantum machines,” Physical Review Applied, vol. 19, no. 2, p. 024027, 2023

    work page 2023

  14. [14]

    Benchmarkin g the performance of portfolio optimization with qaoa,

    S. Brandhofer, D. Braun, V . Dehn, G. Hellstern, M. H¨ uls, Y . Ji, I. Polian, A. S. Bhatia, and T. Wellens, “Benchmarkin g the performance of portfolio optimization with qaoa,” Quantum Information Processing, vol. 22, no. 1, p. 25, 2022

    work page 2022

  15. [15]

    Quantum approximate optimization algorithm: Performance , mechanism, and implementation on near-term devices,

    L. Zhou, S. T. Wang, S. Choi, H. Pichler, and M. D. Lukin, “ Quantum approximate optimization algorithm: Performance , mechanism, and implementation on near-term devices,” Physical Review X, vol. 10, no. 2, p. 021067, 2020

    work page 2020

  16. [16]

    Quantum computing in the nisq era and beyo nd,

    J. Preskill, “Quantum computing in the nisq era and beyo nd,” Quantum, vol. 2, p. 79, 2018

    work page 2018

  17. [17]

    Qaoa for max-cut re quires hundreds of qubits for quantum speed-up,

    G. G. Guerreschi and A. Y . Matsuura, “Qaoa for max-cut re quires hundreds of qubits for quantum speed-up,” Scientific reports, vol. 9, no. 1, p. 6903, 2019

    work page 2019

  18. [18]

    Policy gradient methods for reinforcement learning: A su rvey,

    J. Yao et al. , “Policy gradient methods for reinforcement learning: A su rvey,” F oundations and Trends in Machine Learning , vol. 14, no. 6, pp. 403–500, 2021

    work page 2021

  19. [19]

    Statistical distance and the geometry of quantum states,

    S. L. Braunstein and C. M. Caves, “Statistical distance and the geometry of quantum states,” Physical Review Letters, vol. 72, no. 22, pp. 3439–3443, 1994

    work page 1994

  20. [20]

    C. W. Helstrom, Quantum Detection and Estimation Theory . New Y ork: Academic Press, 1976

    work page 1976

  21. [21]

    Quantum estimation for quantum technolog y,

    M. G. Paris, “Quantum estimation for quantum technolog y,” International Journal of Quantum Information , vol. 7, no. supp01, pp. 125–137, 2009

    work page 2009

  22. [22]

    Quantum metrology from a qua ntum information science perspective,

    G. T´ oth and I. Apellaniz, “Quantum metrology from a qua ntum information science perspective,” Journal of Physics A: Mathematical and Theoretical, vol. 47, no. 42, p. 424006, 2014

    work page 2014

  23. [23]

    Quantum metrology: Fundamental aspects and recent pr ogress,

    R. Demkowicz-Dobrza´ nski, M. Jarzyna, and J. Kołody´ nski, “Quantum metrology: Fundamental aspects and recent pr ogress,” Progress in Optics, vol. 60, pp. 345–435, 2015

    work page 2015

  24. [24]

    V ariational quantum algorithm for estimating the quantum fisher info rmation,

    J. Beckey, M. Cerezo, A. Sone, and P . J. Coles, “V ariational quantum algorithm for estimating the quantum fisher info rmation,” Physical Review Research, 2020

    work page 2020

  25. [25]

    V ariational quantum algorithms to estimate rank, quantum entropies, fidelity, and fisher i nformation via purity minimization,

    K. C. Tan and T. V olkoff, “V ariational quantum algorithms to estimate rank, quantum entropies, fidelity, and fisher i nformation via purity minimization,” Physical Review Research, 2021

    work page 2021

  26. [26]

    Predicting quantum le arnability from landscape fluctuation,

    H.-K. Zhang, C. Zhu, and X. Wang, “Predicting quantum le arnability from landscape fluctuation,” arXiv preprint arXiv:2406.11805 , 2024

    work page arXiv 2024

  27. [27]

    Quantu m natural gradient,

    J. Stokes, J. Izaac, N. Killoran, and G. Carleo, “Quantu m natural gradient,” Quantum, vol. 4, p. 269, 2020

    work page 2020

  28. [28]

    Noise-resilient variational hybrid quantum-c lassical optimization,

    L. Gentini, A. Cuccoli, S. Pirandola, P . V errucchi, and L. Banchi, “Noise-resilient variational hybrid quantum-c lassical optimization,” Physical Review A, 2019

    work page 2019

  29. [29]

    Quantu m metrology assisted by machine learning,

    J. Huang, M. Zhuang, J. Zhou, Y . Shen, and C. Lee, “Quantu m metrology assisted by machine learning,” Advanced Quantum Technologies, p. 2300329, 2024

    work page 2024

  30. [30]

    Quantum natural gradient generalized to noisy and nonunitary circuits,

    B. Koczor and S. C. Benjamin, “Quantum natural gradient generalized to noisy and nonunitary circuits,” Physical Review A , vol. 106, no. 6, p. 062416, 2022

    work page 2022

  31. [31]

    Fisher information in noisy intermediate -scale quantum applications,

    J. J. Meyer, “Fisher information in noisy intermediate -scale quantum applications,” Quantum, vol. 5, p. 539, 2021. 14

    work page 2021

  32. [32]

    Cost function dependent barren plateaus in shallow quantum neural networks,

    M. Cerezo, A. Sone, T. V olkoff, L. Cincio, and P . J. Coles, “Cost function dependent barren plateaus in shallow quantum neural networks,” Nature Communications, vol. 12, no. 1, p. 1791, 2021

    work page 2021

  33. [33]

    Theory of quantum-assisted genetic algorithms,

    M. Larocca, M. Cerezo, K. Sharma, A. Sornborger, and P . J . Coles, “Theory of quantum-assisted genetic algorithms,” Quantum, vol. 6, p. 824, 2022

    work page 2022

  34. [34]

    Heisenberg limit beyond quantum fisher information,

    W. G´ orecki, “Heisenberg limit beyond quantum fisher information,” arXiv preprint arXiv:2304.14370, 2023

    work page arXiv 2023

  35. [35]

    Entanglement detection via quantum fisher information in a coupled atom-field system,

    S. S. Mirkhalaf et al., “Entanglement detection via quantum fisher information in a coupled atom-field system,” Firenze Thesis Repository, 2016

    work page 2016

  36. [36]

    Molecu lar entanglement witness by absorption spectroscopy in cav ity qed,

    W. Wu, F. Fassioli, D. A. Huse, and G. D. Scholes, “Molecu lar entanglement witness by absorption spectroscopy in cav ity qed,” The Journal of Physical Chemistry Letters , vol. 16, pp. 7369–7375, 2024

    work page 2024

  37. [37]

    Quantum metrology with an n-qubit w sup erposition state under noninteracting and interacting ope rations,

    Y . Li and Z. Ren, “Quantum metrology with an n-qubit w sup erposition state under noninteracting and interacting ope rations,” Physical Review A, vol. 107, no. 1, p. 012403, 2023

    work page 2023

  38. [38]

    Estimation of the quantum fisher informat ion on a quantum processor,

    V . Vitale, A. Rath, P . Jurcevic, A. Elben, C. Branciard, and B. V ermersch, “Estimation of the quantum fisher informat ion on a quantum processor,” PRX Quantum, vol. 5, no. 3, p. 030338, 2024

    work page 2024

  39. [39]

    At the limits of criticality-based quantum metro logy: Apparent super-heisenberg scaling revisited,

    M. M. Rams, P . Sierant, O. Dutta, P . Horodecki, and J. Zak rzewski, “At the limits of criticality-based quantum metro logy: Apparent super-heisenberg scaling revisited,” Physical Review X, vol. 8, no. 2, p. 021022, 2018

    work page 2018

  40. [40]

    “super-heisenberg

    Z. Hou, Y . Jin, H. Chen, J.-F. Tang, C.-J. Huang, H. Y uan, G.-Y . Xiang, C.-F. Li, and G.-C. Guo, ““super-heisenberg” a nd heisenberg scalings achieved simultaneously in the estimation of a rot ating field,” Physical Review Letters, vol. 126, no. 7, p. 070503, 2021

    work page 2021

  41. [41]

    Interaction-based quantum metrology showing scaling beyond the heisenberg limit,

    M. Napolitano, M. Koschorreck, B. Dubost, N. Behbood, R . Sewell, and M. W. Mitchell, “Interaction-based quantum metrology showing scaling beyond the heisenberg limit,” Nature, vol. 471, no. 7339, pp. 486–489, 2011

    work page 2011

  42. [42]

    Qaoa for max-cut: A primer,

    G. Guerreschi et al., “Qaoa for max-cut: A primer,” Quantum Science and Technology, vol. 4, no. 2, p. 024002, 2019

    work page 2019

  43. [43]

    Maximum cut problem, max-cut

    C. W. Commander, “Maximum cut problem, max-cut.” Encyclopedia of Optimization , vol. 2, 2009

    work page 2009

  44. [44]

    Solving large scale max cut problems via tabu search,

    G. A. Kochenberger, J.-K. Hao, Z. L¨ u, H. Wang, and F. Glo ver, “Solving large scale max cut problems via tabu search,” Journal of Heuristics, vol. 19, pp. 565–571, 2013

    work page 2013

  45. [45]

    Hybridizing the cro ss-entropy method: An application to the max-cut problem,

    M. Laguna, A. Duarte, and R. Marti, “Hybridizing the cro ss-entropy method: An application to the max-cut problem,” Computers & Operations Research, vol. 36, no. 2, pp. 487–498, 2009. VI. DATA A V AILABILITY The code and QFI results for the problems studied are publicly available at https://github.com/BrianSarmina/Papers/tree/m The QFI matrices used in th...

    work page 2009