pith. sign in

arxiv: 2606.10432 · v1 · pith:FFTWRSYBnew · submitted 2026-06-09 · 🪐 quant-ph

Experimental implementation of continuous-variable QAOA on a quad-rail lattice cluster state

Shota Yokoyama , Atsushi Sakaguchi , Jun-ichi Yoshikawa , Hironari Nagayoshi , Warit Asavanant , Kan Takase , Takuji Hiraoka , Akira Furusawa
show 1 more author
Hidehiro Yonezawa
This is my paper

Pith reviewed 2026-06-27 13:25 UTC · model grok-4.3

classification 🪐 quant-ph
keywords continuous-variable QAOAquad-rail latticecluster statemeasurement-based quantum computingvariational quantum algorithmsquadratic optimizationBayesian optimization
0
0 comments X

The pith

Continuous-variable QAOA is experimentally demonstrated on a quad-rail lattice cluster state for multi-variable quadratic problems.

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

The authors implement the continuous-variable quantum approximate optimization algorithm using a measurement-based platform built on a quad-rail lattice cluster state. They develop a mapping that converts arbitrary quadratic cost functions into operations on this architecture and test it experimentally for problems involving up to 100 modes across several QAOA depths. Performance improves when depth increases from one to two layers, but further depth increases produce only marginal gains in the lab. Idealized simulations assuming unlimited shots and gradient optimization indicate that deeper circuits should continue to improve, pointing to experimental noise and optimizer choice as the current bottlenecks.

Core claim

We experimentally demonstrate the continuous-variable quantum approximate optimization algorithm (CV-QAOA) for multi-variable problems and multiple QAOA depths using a measurement-based CV quantum computing platform on a quad-rail lattice (QRL) cluster state. We propose a systematic method to map arbitrary quadratic cost functions onto the QRL architecture and examine the resulting construction in settings involving up to 100 modes.

What carries the argument

A systematic mapping of arbitrary quadratic cost functions onto the quad-rail lattice cluster state that enables preparation of the CV-QAOA ansatz via programmable measurements.

If this is right

  • Depth-2 CV-QAOA outperforms depth-1 on the tested quadratic problems when run on the QRL platform.
  • Further depth increases produce only limited experimental gains once noise and optimizer limitations dominate.
  • Idealized numerical simulations show that CV-QAOA performance can keep improving with depth when noise and shot limits are removed.
  • The work supplies an explicit mapping and experimental protocol that future CV variational algorithms can build upon.

Where Pith is reading between the lines

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

  • The mapping itself appears scalable to larger mode counts without inherent architectural blockage.
  • Measurement-based CV platforms may require new error-mitigation techniques before deeper QAOA layers become useful.
  • Switching to gradient estimators that work with finite shots could close part of the gap between experiment and ideal simulation.

Load-bearing premise

The observed saturation in performance beyond depth 2 arises mainly from noise accumulation and the choice of classical optimizer rather than from the mapping method or the underlying platform.

What would settle it

An experiment on the same hardware that replaces Bayesian optimization with gradient-based optimization and collects effectively infinite shots to test whether performance continues to rise with depth.

Figures

Figures reproduced from arXiv: 2606.10432 by Akira Furusawa, Atsushi Sakaguchi, Hidehiro Yonezawa, Hironari Nagayoshi, Jun-ichi Yoshikawa, Kan Takase, Shota Yokoyama, Takuji Hiraoka, Warit Asavanant.

Figure 1
Figure 1. Figure 1: FIG. 1. Schematic of the CV-QAOA algorithm. The depth [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. CV-QAOA implementation on the QRL cluster state. (a) Example of the CV-QAOA ansatz mapped onto the [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Cost function landscape and BO process for the two-variable problem at depth [PITH_FULL_IMAGE:figures/full_fig_p006_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Optimization results obtained by BO as a function [PITH_FULL_IMAGE:figures/full_fig_p007_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: FIG. 5. Optimization results for randomly generated quadratic problems with 10 and 100 variables. The functions [PITH_FULL_IMAGE:figures/full_fig_p008_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: FIG. 6. Performance of CV-QAOA evaluated under the idealized conditions using gradient-based optimization with an infinite [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: FIG. 7. QRL graph representation of the experimentally implemented CV-QAOA ansatz for 10-variable quadratic problems [PITH_FULL_IMAGE:figures/full_fig_p011_7.png] view at source ↗
Figure 8
Figure 8. Figure 8: FIG. 8. QRL graph representation of the experimentally im [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
read the original abstract

We experimentally demonstrate the continuous-variable quantum approximate optimization algorithm (CV-QAOA) for multi-variable problems and multiple QAOA depths using a measurement-based CV quantum computing platform on a quad-rail lattice (QRL) cluster state. We propose a systematic method to map arbitrary quadratic cost functions onto the QRL architecture and examine the resulting construction in settings involving up to 100 modes. Using the programmable platform, we prepare the CV-QAOA ansatz and optimize the variational parameters via Bayesian optimization. We then investigate the performance on quadratic optimization problems and observe that increasing the depth from 1 to 2 improves performance, whereas further increases yield only limited gains. In contrast, numerical simulations under idealized conditions, assuming an infinite number of measurement shots and gradient-based optimization, indicate that the performance of CV-QAOA can improve with increasing depth, suggesting that the experimentally observed limitations primarily arise from noise accumulation and classical optimization challenges. This work provides an experimental demonstration of CV-QAOA on a programmable CV platform and establishes a foundation for future developments of variational quantum algorithms in CV systems.

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

1 major / 1 minor

Summary. The manuscript reports an experimental demonstration of continuous-variable QAOA (CV-QAOA) on a programmable measurement-based platform using a quad-rail lattice (QRL) cluster state. It introduces a systematic mapping of arbitrary quadratic cost functions onto the QRL architecture (tested up to 100 modes), prepares the CV-QAOA ansatz for multiple depths, optimizes variational parameters via Bayesian optimization, and observes that performance on quadratic problems improves from depth p=1 to p=2 but saturates thereafter. Idealized numerical simulations (infinite shots, gradient descent, no noise) are shown to improve with depth, leading to the conclusion that experimental saturation arises primarily from noise accumulation and classical optimization difficulties.

Significance. If the attribution of performance limits holds after additional controls, the work constitutes a notable experimental milestone as one of the first realizations of CV-QAOA on a scalable CV platform. The QRL mapping procedure and the direct comparison between hardware and idealized simulations are concrete contributions that could guide future variational CV algorithms.

major comments (1)
  1. [Abstract and Results] Abstract and Results section: The claim that observed saturation beyond depth 2 'primarily arise from noise accumulation and classical optimization challenges' rests on a comparison between hardware data (Bayesian optimization, finite shots, real noise) and idealized simulations (gradient-based optimization, infinite shots, no noise). No intermediate simulations are reported that apply the identical optimizer and shot budget to the QRL-mapped cost function; therefore the data cannot yet distinguish intrinsic limitations of the quad-rail lattice encoding from platform imperfections.
minor comments (1)
  1. [Experimental Methods] The manuscript lacks reported error bars on the experimental performance metrics, raw shot data, and a complete description of the measurement protocol and classical post-processing; these omissions reduce the ability to assess statistical significance of the depth-1 to depth-2 improvement.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the constructive feedback on our manuscript. We address the major comment below and have revised the manuscript to strengthen the comparison between experiment and simulation.

read point-by-point responses

  1. Referee: [Abstract and Results] Abstract and Results section: The claim that observed saturation beyond depth 2 'primarily arise from noise accumulation and classical optimization challenges' rests on a comparison between hardware data (Bayesian optimization, finite shots, real noise) and idealized simulations (gradient-based optimization, infinite shots, no noise). No intermediate simulations are reported that apply the identical optimizer and shot budget to the QRL-mapped cost function; therefore the data cannot yet distinguish intrinsic limitations of the quad-rail lattice encoding from platform imperfections.

    Authors: We agree that the original comparison is not fully controlled because it mixes different optimizers and shot budgets. To resolve this, we have added new idealized simulations that apply the identical Bayesian optimizer and finite-shot budget to the QRL-mapped quadratic cost functions. These simulations show continued improvement from depth 1 to depth 2 (and modest further gains at depth 3), consistent with the experimental trend up to depth 2 but without the saturation seen on hardware. This supports the attribution of experimental saturation to noise and classical optimization difficulties rather than an intrinsic limitation of the encoding. We have updated the abstract and Results section to present these controlled simulations and revised the discussion accordingly. revision: yes

Circularity Check

0 steps flagged

No significant circularity; experimental claims rest on direct measurements and independent simulations

full rationale

The paper reports experimental preparation and optimization of CV-QAOA ansatze on a quad-rail lattice cluster state, with performance evaluated at depths 1-2+ via Bayesian optimization on real hardware. These results are contrasted against separate idealized numerical simulations (gradient descent, infinite shots, no noise). No load-bearing step reduces by construction to a fitted parameter, self-citation, or ansatz smuggled from prior work; the mapping procedure and performance saturation inference are presented as inferences from external benchmarks rather than tautological redefinitions. The derivation chain is self-contained against the reported experimental data and simulations.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The work relies on established principles of continuous-variable quantum optics and measurement-based quantum computation without introducing new fundamental entities or ad-hoc postulates beyond standard domain assumptions.

free parameters (1)
  • QAOA variational parameters
    These parameters are optimized instance-by-instance via Bayesian optimization on the physical device.
axioms (2)
  • standard math Measurement-based quantum computation can be realized on quad-rail lattice cluster states in continuous variables
    Invoked as the foundational resource state for the CV-QAOA ansatz preparation.
  • domain assumption Arbitrary quadratic cost functions admit a systematic mapping onto the QRL architecture
    Stated as the proposed method enabling the experimental tests.

pith-pipeline@v0.9.1-grok · 5758 in / 1323 out tokens · 33895 ms · 2026-06-27T13:25:48.776662+00:00 · methodology

discussion (0)

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

Reference graph

Works this paper leans on

42 extracted references · 5 canonical work pages · 3 internal anchors

  1. [1]

    4, we usen init = 10,n trial = 100, and nattempt = 100 for depthd= 1, andn init = 30,n trial = 200, andn attempt = 20 for depthsd >1

    For Fig. 4, we usen init = 10,n trial = 100, and nattempt = 100 for depthd= 1, andn init = 30,n trial = 200, andn attempt = 20 for depthsd >1. For Fig. 5, we usen init = 10,n trial = 100, andn attempt = 20 for depth d= 1, andn init = 30,n trial = 200, andn attempt = 20 for depthsd >1. Appendix D: Noise analysis of QAOA on the QRL cluster state In this sec...

  2. [2]

    M. A. Nielsen and I. L. Chuang,Quantum Computa- tion and Quantum Information: 10th Anniversary Edi- tion(Cambridge University Press, 2010)

    2010

  3. [3]

    Biamonte, P

    J. Biamonte, P. Wittek, N. Pancotti, P. Rebentrost, N. Wiebe, and S. Lloyd, Quantum machine learning, Na- ture549, 195 (2017)

    2017

  4. [4]

    Y. Cao, J. Romero, J. P. Olson, M. Degroote, P. D. John- son, M. Kieferov´ a, I. D. Kivlichan, T. Menke, B. Per- opadre, N. P. Sawaya,et al., Quantum chemistry in the age of quantum computing, Chemical reviews119, 10856 (2019)

    2019

  5. [5]

    Or´ us, S

    R. Or´ us, S. Mugel, and E. Lizaso, Quantum computing for finance: Overview and prospects, Reviews in Physics 4, 100028 (2019)

    2019

  6. [6]

    Arute, K

    F. Arute, K. Arya, R. Babbush, D. Bacon, J. C. Bardin, R. Barends, R. Biswas, S. Boixo, F. G. Brandao, D. A. Buell,et al., Quantum supremacy using a programmable superconducting processor, Nature574, 505 (2019)

    2019

  7. [7]

    Monroe, W

    C. Monroe, W. C. Campbell, L.-M. Duan, Z.-X. Gong, A. V. Gorshkov, P. W. Hess, R. Islam, K. Kim, N. M. Linke, G. Pagano,et al., Programmable quantum sim- ulations of spin systems with trapped ions, Reviews of Modern Physics93, 025001 (2021)

    2021

  8. [8]

    Ebadi, T

    S. Ebadi, T. T. Wang, H. Levine, A. Keesling, G. Se- meghini, A. Omran, D. Bluvstein, R. Samajdar, H. Pich- ler, W. W. Ho,et al., Quantum phases of matter on a 256-atom programmable quantum simulator, Nature 595, 227 (2021)

    2021

  9. [9]

    Zhong, H

    H.-S. Zhong, H. Wang, Y.-H. Deng, M.-C. Chen, L.-C. Peng, Y.-H. Luo, J. Qin, D. Wu, X. Ding, Y. Hu,et al., Quantum computational advantage using photons, Sci- ence370, 1460 (2020)

    2020

  10. [10]

    Preskill, Quantum computing in the nisq era and be- yond, Quantum2, 79 (2018)

    J. Preskill, Quantum computing in the nisq era and be- yond, Quantum2, 79 (2018)

    2018

  11. [11]

    Cerezo, A

    M. Cerezo, A. Arrasmith, R. Babbush, S. C. Benjamin, S. Endo, K. Fujii, J. R. McClean, K. Mitarai, X. Yuan, L. Cincio,et al., Variational quantum algorithms, Nature Reviews Physics3, 625 (2021)

    2021

  12. [12]

    Kandala, A

    A. Kandala, A. Mezzacapo, K. Temme, M. Takita, M. Brink, J. M. Chow, and J. M. Gambetta, Hardware- efficient variational quantum eigensolver for small molecules and quantum magnets, Nature549, 242 (2017)

    2017

  13. [13]

    A Quantum Approximate Optimization Algorithm

    E. Farhi, J. Goldstone, and S. Gutmann, A quan- tum approximate optimization algorithm, arXiv preprint arXiv:1411.4028 (2014)

    work page internal anchor Pith review Pith/arXiv arXiv 2014

  14. [14]

    M. P. Harrigan, K. J. Sung, M. Neeley, K. J. Satzinger, F. Arute, K. Arya, J. Atalaya, J. C. Bardin, R. Barends, S. Boixo,et al., Quantum approximate optimization of non-planar graph problems on a planar superconducting processor, Nature Physics17, 332 (2021)

    2021

  15. [15]

    Hodson, Brendan Ruck, Hugh Ong, David Garvin, and Stefan Dulman

    M. Hodson, B. Ruck, H. Ong, D. Garvin, and S. Dulman, Portfolio rebalancing experiments using the quantum alternating operator ansatz, arXiv preprint arXiv:1911.05296 (2019)

    work page arXiv 1911

  16. [16]

    Blekos, D

    K. Blekos, D. Brand, A. Ceschini, C.-H. Chou, R.-H. Li, K. Pandya, and A. Summer, A review on quan- tum approximate optimization algorithm and its vari- ants, Physics Reports1068, 1 (2024)

    2024

  17. [17]

    Zhou, S.-T

    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 X10, 021067 (2020)

    2020

  18. [18]

    Larocca, S

    M. Larocca, S. Thanasilp, S. Wang, K. Sharma, J. Bia- monte, P. J. Coles, L. Cincio, J. R. McClean, Z. Holmes, and M. Cerezo, Barren plateaus in variational quantum computing, Nature Reviews Physics7, 174 (2025)

    2025

  19. [19]

    Herrman, P

    R. Herrman, P. C. Lotshaw, J. Ostrowski, T. S. Humble, and G. Siopsis, Multi-angle quantum approximate opti- mization algorithm, Scientific Reports12, 6781 (2022)

    2022

  20. [20]

    A. B. Magann, K. M. Rudinger, M. D. Grace, and M. Sarovar, Feedback-based quantum optimization, Physical Review Letters129, 250502 (2022)

    2022

  21. [21]

    A Quantum Approximate Optimization Algorithm for continuous problems

    G. Verdon, J. M. Arrazola, K. Br´ adler, and N. Killoran, A quantum approximate optimization algorithm for contin- uous problems, arXiv preprint arXiv:1902.00409 (2019)

    work page internal anchor Pith review Pith/arXiv arXiv 1902

  22. [22]

    Asavanant and A

    W. Asavanant and A. Furusawa,Optical quantum com- puters: A Route to Practical Continuous Variable Quantum Information Processing(AIP Publishing LLC, 2022)

    2022

  23. [23]

    Asavanant, Y

    W. Asavanant, Y. Shiozawa, S. Yokoyama, B. Charoen- sombutamon, H. Emura, R. N. Alexander, S. Takeda, J.-i. Yoshikawa, N. C. Menicucci, H. Yonezawa,et al., Generation of time-domain-multiplexed two-dimensional cluster state, Science366, 373 (2019)

    2019

  24. [24]

    M. V. Larsen, X. Guo, C. R. Breum, J. S. Neergaard- Nielsen, and U. L. Andersen, Deterministic generation of a two-dimensional cluster state, Science366, 369 (2019)

    2019

  25. [25]

    M. V. Larsen, X. Guo, C. R. Breum, J. S. Neergaard- Nielsen, and U. L. Andersen, Deterministic multi-mode gates on a scalable photonic quantum computing plat- form, Nature Physics17, 1018 (2021)

    2021

  26. [26]

    A full-stack analog optical quantum computing platform with one hundred inputs

    S. Yokoyama, A. Sakaguchi, W. Asavanant, K. Takase, Y.-R. Chen, H. Nagayoshi, J.-i. Yoshikawa, T. Kashi- wazaki, A. Inoue, T. Umeki,et al., Full-stack analog op- tical quantum computer with a hundred inputs, arXiv preprint arXiv:2506.16147 (2025)

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  27. [27]

    Enomoto, K

    Y. Enomoto, K. Anai, K. Udagawa, and S. Takeda, Continuous-variable quantum approximate optimization on a programmable photonic quantum processor, Physi- cal Review Research5, 043005 (2023)

    2023

  28. [28]

    N. C. Menicucci, Temporal-mode continuous-variable cluster states using linear optics, Physical Review A83, 062314 (2011)

    2011

  29. [29]

    R. N. Alexander and N. C. Menicucci, Flexible quantum circuits using scalable continuous-variable cluster states, Physical Review A93, 062326 (2016)

    2016

  30. [30]

    Yoshikawa, W

    J.-i. Yoshikawa, W. Asavanant, H. Nagayoshi, A. Sak- aguchi, S. Yokoyama, H. Yonezawa, and A. Furusawa, Configuration design of multimode gaussian operations on continuous-variable quad-rail lattice cluster states, arXiv preprint arXiv:2506.11236 (2025)

    work page arXiv 2025

  31. [31]

    The RIKEN Optical Quantum Computing Project, mqc3: An open-source python sdk for the measurement- based quantum computer with continuous-variable clus- ter states (2025), available athttps://github.com/ HidehiroYonezawa/mqc3

    2025

  32. [32]

    Shahriari, K

    B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, and N. De Freitas, Taking the human out of the loop: A review of bayesian optimization, Proceedings of the IEEE 104, 148 (2015). 14

    2015

  33. [33]

    J. A. Nielsen, M. J. Kicinski, T. N. Arge, K. Vijayad- haran, J. Foldager, J. Borregaard, J. J. Meyer, J. S. Neergaard-Nielsen, T. Gehring, and U. L. Andersen, Variational quantum algorithm for enhanced continuous variable optical phase sensing, npj Quantum Information 11, 70 (2025)

    2025

  34. [34]

    Akiba, S

    T. Akiba, S. Sano, T. Yanase, T. Ohta, and M. Koyama, Optuna: A next-generation hyperparameter optimiza- tion framework, inProceedings of the 25th ACM SIGKDD international conference on knowledge discov- ery & data mining(2019) pp. 2623–2631

    2019

  35. [35]

    Akshay, D

    V. Akshay, D. Rabinovich, E. Campos, and J. Biamonte, Parameter concentrations in quantum approximate opti- mization, Physical Review A104, L010401 (2021)

    2021

  36. [36]

    Z. He, R. Shaydulin, S. Chakrabarti, D. Herman, C. Li, Y. Sun, and M. Pistoia, Alignment between initial state and mixer improves qaoa performance for constrained op- timization, npj Quantum Information9, 121 (2023)

    2023

  37. [37]

    S. H. Sack and M. Serbyn, Quantum annealing initializa- tion of the quantum approximate optimization algorithm, quantum5, 491 (2021)

    2021

  38. [38]

    Campos, D

    E. Campos, D. Rabinovich, V. Akshay, and J. Biamonte, Training saturation in layerwise quantum approximate optimization, Physical Review A104, L030401 (2021)

    2021

  39. [39]

    Schuld, V

    M. Schuld, V. Bergholm, C. Gogolin, J. Izaac, and N. Kil- loran, Evaluating analytic gradients on quantum hard- ware, Physical Review A99, 032331 (2019)

    2019

  40. [40]

    Lloyd and S

    S. Lloyd and S. L. Braunstein, Quantum computation over continuous variables, Physical Review Letters82, 1784 (1999)

    1999

  41. [41]

    Gottesman, A

    D. Gottesman, A. Kitaev, and J. Preskill, Encoding a qubit in an oscillator, Physical Review A64, 012310 (2001)

    2001

  42. [42]

    Sakaguchi, S

    A. Sakaguchi, S. Konno, F. Hanamura, W. Asavanant, K. Takase, H. Ogawa, P. Marek, R. Filip, J.-i. Yoshikawa, E. Huntington,et al., Nonlinear feedforward enabling quantum computation, Nature Communications14, 3817 (2023)

    2023