Pith. sign in

REVIEW 4 major objections 5 minor 1 cited by

Topological network analysis using a programmable photonic quantum processor

T0 review · 4 major / 5 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 A programmable photonic quantum processor can find heavy cliques, estimate Betti numbers, and flag percolation transitions in weighted networks.

desk verdict A real experimental proof-of-principle for GBS-based network analysis, but the Betti-number estimate is unvalidated and needs exact comparison before the central claim can stand. read the letter →

arxiv 2507.08157 v2 pith:QFR4B6FH submitted 2025-07-10 quant-ph math-phmath.MP

Shang Yu , Jinzhao Sun , Zhenghao Li , Ewan Mer , Yazeed K Alwehaibi , Oscar Scholin , Gerard J. Machado , Kuan-Cheng Chen
show 6 more authors
Aonan Zhang Raj B Patel Ying Dong Ian A. Walmsley Vlatko Vedral Ginestra Bianconi
This is my paper · ORCID
classification quant-phmath-phmath.MP
keywords GaussianbosonsamplingtopologicaldataanalysisBettinumbersk-cliqueidentificationcomplex-weightednetworkscliquepercolationRényientropyprogrammablephotonicprocessor
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 claims that Gaussian boson sampling on a programmable photonic processor can serve as a practical route into topological data analysis for complex-weighted networks. Because the sampling distribution preferentially produces high-weight, densely connected subgraphs, the photon patterns surface the k-cliques most relevant to the network's structure before a classical post-processing step. From those cliques the authors build boundary matrices, estimate Betti numbers, track the Euler characteristic through a two-dimensional filtering process, and locate topological phase transitions. They further find that the normalized Rényi entropy of the sampling patterns closely tracks the clique-percolation order parameter, offering an entropy-based percolation indicator. The central contribution is a new application: photonic quantum sampling as a heuristic guide for the computationally hard step of finding the simplicial complex underlying a network.

What carries the argument

The central machinery is Gaussian boson sampling (GBS), a photonic scheme in which squeezed light passes through a reconfigurable interferometer and photon-number measurements sample subgraphs with probability biased toward high edge weights and dense connectivity. The network's adjacency matrix enters through the decomposition $A' = U \oplus_i \tanh(r_i) U^T$, so an arbitrary complex-weighted graph can be programmed into the device. That bias is what turns sampling patterns into a guide to weighted k-cliques; the boundary matrices built from those cliques carry the Betti-number computation, and the Rényi entropy of the sampling patterns carries the percolation signal.

What would settle it

On a small weighted network whose exact Betti numbers and percolation threshold are known from exhaustive clique enumeration, one could encode the network in the photonic processor and compare the GBS-derived $\beta_k$ and $\tilde{H}_2$-based threshold with the exact values while varying the sampling-probability threshold and the number of samples. If the estimates converge to the topology of a sampling-selected subcomplex rather than to the input network's true topology, the representativeness assumption fails.

Watch

Extended reading notes

Core claim

On a reconfigurable photonic processor that encodes complex-valued adjacency matrices into squeezed-light interference, the paper claims that the output statistics of Gaussian boson sampling are biased toward high-weight dense subgraphs. The sampled k-cliques, filtered by a sampling-probability threshold and refined by a classical greedy-shrinking search, are used to reconstruct the network's clique complex; Betti numbers follow from the boundary-matrix formula $\beta_k = m_k - r_k - r_{k+1}$, and the Euler characteristic $\chi = \sum_k (-1)^k m_k$ locates topological phase transitions where $\chi$ vanishes. The paper also reports that the normalized Rényi entropy $\tilde{H}_2$ of k-photon sampling patterns closely mirrors the clique-percolation order parameter $\Phi$, and that shifts in this entropy track deliberately engineered topological damage. The stated objective is not a quantum speedup for Betti-number estimation but a sampling-based method that returns the identities and weights of k-cliques at once, which the authors describe as an avenue not previously explored.

Load-bearing premise

The argument depends on the assumption that the high-weight cliques GBS preferentially samples are the ones that determine a network's topological structure, so that cutting the clique complex at a sampling-probability threshold yields Betti numbers and percolation signals that describe the original network rather than a sampling-biased subcomplex.

Editorial extensions

If this is right

  • For weighted networks with uneven weight distributions, the GBS bias narrows the search to high-weight regions first, so the method can find high-weight k-cliques faster than uniform or squashed-state classical sampling baselines.
  • The same sampled cliques feed a boundary-matrix construction that yields Betti numbers and Euler characteristics, giving a clique-density-guided filtering process for weighted networks.
  • The normalized Rényi entropy of k-photon sampling patterns can serve as a percolation indicator, identifying percolation thresholds and topological damage without enumerating all k-cliques.
  • The processor is programmable and modular, and the paper presents it as scalable toward hundreds of modes, with applications envisaged in brain-network analysis, protein engineering, and connectivity in materials or porous media.
  • The paper explicitly stops short of claiming quantum computational advantage for Betti-number estimation; the claimed contribution is the sampling-based construction of the simplicial complex, not a complexity-theoretic speedup.

Reading between the lines

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

  • The authors leave implicit that the sampling-probability threshold defines a measurement-selected subcomplex, so the reported Betti numbers are best read as homology estimates of that subcomplex; a direct comparison with exact Betti numbers on small synthetic networks would make this explicit.
  • A testable extension is to run the Rényi-entropy percolation indicator online while a network's weights drift, using it as a dynamic order parameter in streaming or evolving networks; the paper hints at dynamic application but does not demonstrate it.
  • One could ablate the imaginary part of the complex edge weights and rerun the experiment to see how much of the topological signal comes from the phase layer rather than the real-valued magnitudes, directly testing the claimed advantage of complex-weight encoding over non-negative graph encodings.
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

4 major / 5 minor

Summary. The paper reports a programmable photonic Gaussian boson sampling (GBS) processor, 'Babbage,' with temporal-polarization-hybrid encoding, and applies it to topological network analysis of a random 40-node dual-layer network with complex edge weights. The authors claim three main results: (i) GBS-assisted heuristic search finds high-weight k-cliques with improved success rate over classical uniform sampling and squashed-state sampling; (ii) Betti numbers of the network can be estimated by thresholding GBS sampling probabilities, filtering cliques, and computing boundary-matrix ranks; and (iii) the normalized Rényi entropy of GBS sampling patterns tracks k-clique percolation and can detect topological phase transitions through the Euler characteristic.

Significance. If validated, the work would demonstrate a practical, near-term use of GBS for network science, especially for complex-weighted graphs where classical dense-subgraph methods are less developed. The paper's strengths include a fully programmable photonic platform, use of the externally established GBS bias toward heavy dense subgraphs, and a genuinely interesting proposal to use sampling entropy as a percolation indicator without exhaustive clique enumeration. However, the central quantitative claims currently rest on a single network realization, small sample counts, and threshold choices derived from the same data used to define the reconstructed complex, with no ground-truth validation for the Betti-number estimates.

major comments (4)
  1. [Methods, 'Boundary matrix and Betti number'; Fig. 2b] The Betti-number pipeline is not validated against exact Betti numbers. The threshold used to select 5-cliques is 'determined by sorting the probability distribution of GBS sampling results,' so the same data that defines the selected clique set also sets the cut. Because GBS is biased toward high-weight dense subgraphs, the reconstructed graph is a sampling-biased subcomplex rather than the input network, and βk = mk − rk − rk+1 computed from it may describe holes and components of that subcomplex. For a 40-node network, exact Betti numbers are classically computable, so the authors should compare their estimates with ground truth across the threshold range and report the discrepancy. Without this comparison, the abstract's concrete deliverable of 'estimate Betti numbers' is unsupported.
  2. [Fig. 1c and Methods, 'Improvement in the success rate of GBS compared to classical sampling'] The claimed clique-search enhancement is demonstrated on one random dual-layer network with approximately 3000 samples and no error bars. The improvement ratio p_GBS/p_classical and p_GBS/p_squashed is a single realization and could be dominated by sampling noise or by the details of the post-processing (greedy shrinking and local search). The authors should report results over multiple random network instances, give standard errors or confidence intervals, and specify whether the identical post-processing was applied to all three input types. Without these, the 'improvement' in Fig. 1c is not statistically established.
  3. [Methods, 'Details about the TPH-encoding photonic quantum processor'] The network encoding uses A′ = cA + dI with rescaling constants c and d chosen to ensure that all eigenvalues tanh r_i lie between 0 and 1. These constants are free parameters that directly control the sampling probability distribution and therefore affect every downstream quantity: clique identification, Betti numbers, Euler characteristics, and Rényi entropy. The manuscript does not specify how c and d are chosen, nor how sensitive the results are to this choice. The authors should provide an explicit prescription and a sensitivity analysis, since otherwise the experimental mapping from network weights to GBS probabilities is underdetermined.
  4. [Fig. 4 and Methods, 'Details about using Rényi entropy to detect clique percolation'] The claim that normalized Rényi entropy H̃2 'closely mirrors' the percolation order parameter Φ is supported only by visual inspection. No quantitative agreement metric (e.g., correlation coefficient, root-mean-square error, or threshold-estimation error) is given, and the 5-photon entropy curve is based on only about 3000 samples, so statistical uncertainty is non-negligible. The order α = 2 is also chosen without justification or robustness testing. The authors should quantify the agreement, provide error bars, and show that the qualitative conclusion is stable under variations of α and sample size.
minor comments (5)
  1. [Main text, Discussion] The sentence 'Although a quantum computational advantage in estimating Betti numbers not exhibit' is missing a verb and should be rewritten (e.g., 'Although we do not exhibit a quantum computational advantage...').
  2. [Main text, Discussion] The phrase 'there is currently no efficient classical method can simulate the outcomes of GBS experiments' should read 'there is currently no efficient classical method that can simulate...'.
  3. [References] Reference [40] is listed as 'Phys. Mon. Phys. 80, 1275-1335 (2008)'; this should be 'Rev. Mod. Phys. 80, 1275-1335 (2008)'.
  4. [References] Reference [62] is cited as 'PRX Quantum 37, 100247 (2020)'; PRX Quantum does not use volume 37, and the title 'Big networks: A survey' suggests the journal and volume are incorrect. Please verify and correct this reference.
  5. [Fig. 1 caption] The caption mentions a '220 m lone-fibre'; this should presumably be '220 m long-fibre'.

Circularity Check

1 steps flagged · score 4.0 of 10

Betti-number 'estimates' are computed from a network reconstructed using a threshold read off the same GBS sampling distribution, so they describe the GBS-selected subcomplex rather than an independently validated property of the input network.

  1. fitted input called prediction [Main text, 'Calculating Betti numbers using Babbage' (Fig. 2b); also Methods, 'Boundary matrix and Betti number'.]
    "In Fig. 2b, we begin by reconstructing the network using the 5-cliques that exceed a certain density threshold, determined by sorting the probability distribution of GBS sampling results, and then compute all the necessary k-clique information for calculating βk."

    The threshold is not an independent filtration parameter derived from the network weights; it is read off the GBS sampling distribution itself. The same distribution's high-probability events select the 5-cliques that define the reconstructed graph, so the resulting Betti numbers are a function of the GBS output by construction. Reporting them as an 'estimate' of the network's Betti numbers is therefore not validated against the input graph; it is a description of the GBS-selected subcomplex. The paper provides no comparison with exact Betti numbers (classically computable for these graph sizes), so the reported βk cannot be distinguished from the selection rule that produced them.

full rationale

The paper's core mechanism—GBS preferentially sampling high-weight, dense subgraphs—is externally established (Refs [35–38]), so using GBS to find heavy k-cliques is not circular. Likewise, the Rényi-entropy/percolation comparison is an empirical correlation between the GBS sampling entropy and a separately computed order parameter Φ, so that result has independent content. The main circular-adjacent step is the Betti-number pipeline: the threshold defining the filtered clique complex is 'determined by sorting the probability distribution of GBS sampling results,' and the same selected 5-cliques are then used to compute βk. Thus βk describes the homology of a subcomplex selected by GBS probabilities, not an independently validated property of the original network. No exact Betti numbers are provided as ground truth, and the paper concedes that complete k-clique search remains challenging and that GBS must be combined with classical heuristics. This is a localized, partial fitted-input issue rather than a self-citation chain; the other main results stand independently.

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

The central mechanism rests on the established GBS dense-subgraph bias, on TDA definitions, and on a new, unvalidated entropy-percolation correspondence. No invented entities are introduced.

free parameters (4)
  • Rescaling constants c,d = not reported
    A' = cA + dI is chosen so all eigenvalues tanh r_i lie in [0,1] for hardware encoding (Methods). The values depend on the input network and are chosen by hand.
  • Sampling probability threshold for Betti filtering = not reported
    In Fig 2b and Methods, cliques are kept if their GBS sampling probability exceeds a threshold set by sorting the probability distribution. This threshold is selected post hoc and directly determines the reconstructed network and resulting Betti numbers.
  • Rényi entropy order alpha = alpha=2
    The paper uses H2 for percolation detection; alpha is chosen rather than derived, and affects how strongly rare events are damped.
  • Euler characteristic filtering thresholds (omega_t, delta_t) = swept; path chosen post hoc
    The two-dimensional filter thresholds are swept and the path in Fig 3a is selected to reveal a topological phase transition; the reported transition point depends on this choice.
assumptions (4)
  • domain assumption GBS sampling probability favors high-weight dense subgraphs.
    Central mechanism, established in Refs 35-38 and used to motivate clique identification.
  • domain assumption The clique complex built from identified k-cliques represents the network's topology.
    Underlies Betti number computation; the paper assumes sampled cliques suffice.
  • domain assumption Euler characteristic vanishing marks topological phase transitions.
    Taken from Ref 41; used to define TPT in Fig 3.
  • ad hoc to paper Rényi entropy of GBS sampling patterns tracks clique percolation.
    This correspondence is proposed and tested here, not established; it is the paper's new heuristic and is not independently validated quantitatively.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Topological network analysis using a programmable photonic quantum processor." pith.science (2026). https://pith.science/paper/QFR4B6FH

@misc{pith2026250708157,
  author       = {Pith},
  title        = {Pith review of: Topological network analysis using a programmable photonic quantum processor},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QFR4B6FH}},
  note         = {Machine review of arXiv:2507.08157}
}
abstract

Understanding topological features in networks is crucial for unravelling complex phenomena across fields such as neuroscience, condensed matter, and high-energy physics. However, identifying higher-order topological structures -- such as $k$-cliques, fundamental building blocks of complex networks -- remains a significant challenge. Here we develop a universal programmable photonic quantum processor that enables the encoding of arbitrary complex-weight networks, providing a direct pathway to uncovering their topological structures. We demonstrate how this quantum approach can identify weighted $k$-cliques and estimate Betti numbers by leveraging the Gaussian boson sampling algorithm's ability to preferentially select high-weight, dense subgraphs. The unique capabilities of our programmable quantum processor allow us to observe topological phase transitions and identify clique percolation phenomena directly from the entropy of the sampling results. These findings showcase how photonic quantum computing can be applied to analyse the topological characteristics of real-world complex networks, opening new possibilities for quantum-enhanced data analysis.

Figures

Figures reproduced from arXiv: 2507.08157 by the authors.

Figure 1
Figure 1. FIG. 1 [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. FIG. 2 [PITH_FULL_IMAGE:figures/full_fig_p004_2.png] view at source ↗
Figure 3
Figure 3. FIG. 3 [PITH_FULL_IMAGE:figures/full_fig_p005_3.png] view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: FIG. 4 [PITH_FULL_IMAGE:figures/full_fig_p006_4.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Gaussian Boson Sampling for Asset Clustering in Statistical Arbitrage Portfolios

    quant-ph 2026-07 conditional novelty 6.0 of 10

    GBS-based clustering (GBS Roots and adapted GBS Boost) produced higher StatArb portfolio returns than classical Spectral/SPONGE clustering in simulated S&P 500 backtests, with the advantage shrinking outside high-vola...

Reference graph

Works this paper leans on

78 extracted references · 76 canonical work pages · cited by 1 Pith paper

  1. [65]

    & Jiang, L

    Oh, C., Liu, M., Alexeev, Y., Fefferman, B. & Jiang, L. Classical algorithm for simulating experimental Gaussian boson sampling. Nat. Phys. 20, 1461-1468 (2024)

    work page 2024

  2. [1]

    Twenty years of network science

    Vespignani, A. Twenty years of network science. Nature 558, 528-529 (2018)

    work page 2018

  3. [2]

    Bianconi, G., Higher-order networks: an introduction to simplicial complexes Cambridge University Press (2021)

    work page 2021

  4. [3]

    & Newman, M

    Fortunato, S. & Newman, M. E. J. 20 years of network community detection. Nat. Phys. 18, 848-850 (2022)

    work page 2022

  5. [4]

    P´ osfai, M. et. al. Impact of physicality on network struc- ture. Nat. Phys. 20, 142-149 (2024). 7

    work page 2024

  6. [5]

    & Sporns O., Complex brain networks: graph theoretical analysis of structural and functional systems

    Bullmore, E. & Sporns O., Complex brain networks: graph theoretical analysis of structural and functional systems. Nat. Rev. Neuroscience110, 186-198 (2009)

    work page 2009

  7. [6]

    S & Sporns, O

    Bassett D. S & Sporns, O. Network neuroscience. Nat. Neurosci. 20, 353-364 (2017)

    work page 2017

  8. [7]

    & Wei, G.-W

    Wang, M., Cang, Z. & Wei, G.-W. A topology-based net- work tree for the prediction of protein-protein binding affinity changes following mutation. Nat. Mach. Intell.2, 116-123 (2020)

    work page 2020

Show all 78 references
  1. [8]

    & Xia, K., Persistent spectral-based machine learning (PerSpect ML) for protein-ligand binding affin- ity prediction

    Meng, Z. & Xia, K., Persistent spectral-based machine learning (PerSpect ML) for protein-ligand binding affin- ity prediction. Sci. Adv. 7, eabc5329 (2021)

    2021

  2. [9]

    & Wei, G., Persistent spectral theory-guided pro- tein engineering, Nat

    Qiu, Y. & Wei, G., Persistent spectral theory-guided pro- tein engineering, Nat. Comput. Sci.3, 149-163 (2023)

    2023

  3. [10]

    & Wei, G., Multiscale topology- enabled structure-to-sequence transformer for protein- ligand interaction predictions

    Chen, D., Liu, J. & Wei, G., Multiscale topology- enabled structure-to-sequence transformer for protein- ligand interaction predictions. Nat. Mach. Intell.6, 799- 810 (2024)

    2024

  4. [11]

    Grewal, D. S. Network power: The social dynamics of globalization. Yale University Press (2008)

    2008

  5. [12]

    W., Papadopoulos, L., Kahn, A

    Lynn, C. W., Papadopoulos, L., Kahn, A. E. & Bas- sett, D. S. Human information processing in complex networks. Nat. Phys. 16, 965-973 (2020)

    2020

  6. [13]

    Alvarez-Rodriguez, U. et. al. Evolutionary dynamics of higher-order interactions in social networks. Nat. Hum. Behav. 5, 586-595 (2021)

    2021

  7. [14]

    A., Carlsson, G

    Baas, N. A., Carlsson, G. E., Quick, G., Szymik, M. & Thaule, M. Topological data analysis. The Abel Sympo- sium 2018, Springer Press (2018)

    2018

  8. [15]

    E., Phillips-Cremins, J

    Sizemore, A. E., Phillips-Cremins, J. E., Ghrist, R. & Bassett, D. S. The importance of the whole: Topologi- cal data analysis for the network neuroscientist. Network Neuroscience 3 656-673 (2019)

    2019

  9. [16]

    & Vejdemo-Johansson, M

    Carlsson, G. & Vejdemo-Johansson, M. Topological data analysis with applications. Cambridge University Press (2022)

    2022

  10. [17]

    and Harrington, H.A

    Otter, N., Porter, M.A., Tillmann, U., Grindrod, P. and Harrington, H.A. A roadmap for the computation of per- sistent homology. EPJ Data Science, 6, 1-38 (2017)

    2017

  11. [18]

    Topological methods for data modelling

    Carlsson, G. Topological methods for data modelling. Nat. Rev. Phys.2, 697-708 (2020)

    2020

  12. [19]

    On bounding the Betti numbers and comput- ing the euler characteristic of semi-algebraic sets.Discret

    Basu, S. On bounding the Betti numbers and comput- ing the euler characteristic of semi-algebraic sets.Discret. Comput. Geom. 22, 1-18 (1999)

    1999

  13. [20]

    Computing the top Betti numbers of semi- algebraic sets defined by quadratic inequalities in poly- nomial time

    Basu, S. Computing the top Betti numbers of semi- algebraic sets defined by quadratic inequalities in poly- nomial time. Found. Comput. Math.8, 45-80 (2008)

    2008

  14. [21]

    Lloyd, S. et. al. Quantum algorithms for topological and geometric analysis of data. Nat. Commun. 7, 1 (2016)

    2016

  15. [22]

    & Lloyd, S

    Schmidhuber, A. & Lloyd, S. Complexity-theoretic limi- tations on quantum algorithms for topological data anal- ysis. PRX Quantum 4, 040349 (2023)

    2023

  16. [23]

    Harrigan, M. P. et al., Quantum approximate optimiza- tion of non-planar graph problems on a planar supercon- ducting processor. Nat. Phys. 17, 332-336 (2021)

    2021

  17. [24]

    Berry, D. W. et. al. Analyzing prospects for quantum advantage in topological data analysis. PRX Quantum 5, 010319 (2024)

    2024

  18. [25]

    Huang, H.-L. et. al. Demonstration of topological data analysis on a quantum processor. Optica 5, 193 (2018)

    2018

  19. [26]

    & Kohler, T., Clique Homology is QMA 1- hard

    Crichigno, M. & Kohler, T., Clique Homology is QMA 1- hard. Nat. Commun. 15, 9846 (2024)

    2024

  20. [27]

    Akhalwaya, L. Y. et. al. Topological data analysis on noisy quantum computers. Preprint at arXiv: 2209.09371 (2024)

    2024 arXiv

  21. [28]

    & Hayakawa, R., Quantum computing and persistence in topological data analysis

    Gyurik, C., Schmidhuber, A., King, R., Dunjko, V. & Hayakawa, R., Quantum computing and persistence in topological data analysis. Preprint at arXiv: 2410.21258 (2024)

    2024 arXiv

  22. [29]

    To- wards quantum advantage via topological data analysis

    Gyurik, Casper, Chris Cade, and Vedran Dunjko. To- wards quantum advantage via topological data analysis. Quantum 6 (2022): 855

    2022

  23. [30]

    Lund, A. P. et al. Boson sampling from a Gaussian state. Phys. Rev. Lett.113, 100502 (2014)

    2014

  24. [31]

    Hamilton, C. S. et al. Gaussian boson sampling. Phys. Rev. Lett.119, 170501 (2017)

    2017

  25. [32]

    Zhong, H.-S. et al. Quantum computational advantage using photons. Science 370, 1460-1463 (2020)

    2020

  26. [33]

    Madsen, L. S. et al. Quantum computational advantage with a programmable photonic processor. Nature 606, 75-81 (2022)

    2022

  27. [34]

    Yu, S. et. al. A universal programmable Gaussian boson sampler for drug discovery. Nat. Comput. Sci.3, 839-848 (2023)

    2023

  28. [35]

    Arrazola, J. M. & Bromley, T. R. Using Gaussian boson sampling to find dense subgraphs. Phys. Rev. Lett.121, 030503 (2018)

    2018

  29. [36]

    B., Walmsley, I

    Sempere-Llagostera, S., Patel, R. B., Walmsley, I. A. & Kolthammer, W. S. Experimentally finding dense sub- graphs using a time-bin encoded Gaussian boson sam- pling device. Phys. Rev. X12, 031045 (2022)

    2022

  30. [37]

    Deng, Y. et. al. Solving graph problems using Gaussian boson sampling. Phys. Rev. Lett.130, 190601 (2023)

    2023

  31. [38]

    Jiang, L

    Oh, C., Fefferman, B. Jiang, L. & Quesada, N. Quantum- inspired classical algorithm for graph problems by Gaus- sian boson sampling. PRX Quantum 5, 020341 (2024)

    2024

  32. [39]

    Boccaletti, S., Latora, V., Moreno, Y., Chavez, M., Hwang, D.-U., Complex networks: Structure and dynam- ics. Phys. Rep. 424, 175-308 (2006)

    2006

  33. [40]

    N., Goltsev, A

    Dorogovtsev, S. N., Goltsev, A. V. & Mendes, J. F. F. Critical phenomena in complex networks. Phys. Mon. Phys. 80, 1275-1335 (2008)

    2008

  34. [41]

    Santos, F. A. N. et al., Topological phase transitions in functional brain networks. Phys. Rev. E 100, 032414 (2019)

    2019

  35. [42]

    & Vicsek, T

    Der´ enyi, I., Palla, G. & Vicsek, T. Clique percolation in random networks. Phys. Rev. Lett.94, 160202 (2005)

    2005

  36. [43]

    S. Yu et. al. Experimental investigation of quantum PT - enhanced sensor. Phys. Rev. Lett.125, 240506 (2020)

    2020

  37. [44]

    R., Humphreys, P

    Clements, W. R., Humphreys, P. C., Metcalf, B. J., Kolthammer, W. S. & Walmsley, I. A. Optimal design for universal multiport interferometers. Optica 12, 1460- 1465 (2016)

    2016

  38. [45]

    L., Neergaard-Nielsen, J

    Andersen, U. L., Neergaard-Nielsen, J. S., van Loock, P. & Furusawa, A. Hybrid discrete- and continuous-variable quantum information. Nat. Phys. 11, 713-719 (2015)

    2015

  39. [46]

    Universal quantum comput- ing with measurement-induced continuous-variable gate sequence in a loop-based architecture

    Takeda, S., & Furusawa, A. Universal quantum comput- ing with measurement-induced continuous-variable gate sequence in a loop-based architecture. Phys. Rev. Lett. 119, 120504 (2017)

    2017

  40. [47]

    Asavanant, W. et. al. Generation of time-domain- multiplexed two-dimensional cluster state. Science 366, 373-376 (2019)

    2019

  41. [48]

    V., Guo, X., Breum, C

    Larsen, M. V., Guo, X., Breum, C. R, Neergaard-Nielsen, J. S, Andersen, Ulrik L, Deterministic generation of a two-dimensional cluster state. Science 366, 369-372 (2019)

    2019

  42. [49]

    Enomoto, Y., Yonezu, K., Mitsuhashi, Y., Takase, K., 8 & Takeda, S., Programmable and sequential Gaussian gatesin a loop-based single-mode photonicquantum pro- cessor. Sci. Adv. 7, eabj6624 (2021)

    2021

  43. [50]

    Time-domain universal linear-optical operations for uni- versal quantum information processing

    Yonezu, K., Enomoto, Y., Yoshida, T., & Takeda, S. Time-domain universal linear-optical operations for uni- versal quantum information processing. Phys. Rev. Lett. 131, 040601 (2023)

    2023

  44. [51]

    Yu, S. et. al. A von-Neumann-like photonic processor and its application in studying quantum signature of chaos. Light Sci. Appl.13, 74 (2024)

    2024

  45. [52]

    A., Complex networks with complex weights

    B¨ ottcher, L., & Porter, M. A., Complex networks with complex weights. Phys. Rev. E109, 024314 (2024)

    2024

  46. [53]

    Paul, C. R. Analysis of Multiconductor Transmission Lines. John Wiley & Sons, Hoboken, NJ, USA, (2007)

    2007

  47. [54]

    Strub, S. H. & B¨ ottcher, L. Modeling deformed trans- mission lines for continuous strain sensing applications, Meas. Sci. Technol.31, 035109 (2020)

    2020

  48. [55]

    Complex quantum networks: a topical review, Phys

    Nokkala, J., Piilo, J & Bianconi, G. Complex quantum networks: a topical review, Phys. A: Math. Theor. 57 233001 (2024)

    2024

  49. [56]

    Kobayashi, Symmetric complex-valued Hopfield neu- ral networks

    M. Kobayashi, Symmetric complex-valued Hopfield neu- ral networks. IEEE Trans. Neural Netw. Learn. Syst.28, 1011 (2016)

    2016

  50. [57]

    Zhang, H. et. al. An optical neural chip for implementing complex-valued neural network. Nat. Commun. 12, 457 (2021)

    2021

  51. [58]

    Spall, J., Guo, X., & Lvovsky, A. I. Hybrid training of optical neural networks, Optica 9, 803 (2022)

    2022

  52. [59]

    Dai, X. K. et. al. D-dimensional oscillators in simplicial structures: Odd and even dimensions display different synchronization scenarios. Chaos, Solitons Fractals146, 803 (2021)

    2021

  53. [60]

    & Arra- zola, J

    Banchi, L., Fingerhuth, M., Babej, T., Ing, C. & Arra- zola, J. M. Molecular docking with Gaussian boson sam- pling. Sci. Adv. 6, eaax1950 (2020)

    2020

  54. [61]

    M., & Que- sada, N

    Mart ´ ınez-Cifuentes, J., Fonseca-Romero, K. M., & Que- sada, N. Classical models may be a better explanation of the Jiuzhang 1.0 Gaussian Boson Sampler than its tar- geted squeezed light model. Quantum 7, 1076 (2023)

    2023

  55. [62]

    Bedru, H. D. Big networks: A survey. PRX Quantum37, 100247 (2020)

    2020

  56. [63]

    Battiston, F. et. al. The physics of higher-order inter- actions in complex systems. Nat. Phys. 17, 1093-1098 (2021)

    2021

  57. [64]

    & Chen, G

    Shi, D., Chen, Z., Sun, X., Chen, Q., Ma, C., Lou, Y. & Chen, G. Computing cliques and cavities in networks. Commun. Phys. 4, 249 (2021)

    2021

  58. [66]

    W., Elementary Applied Topology (CreateS- pace, Scotts Valley, CA, 2016)

    Ghrist, R. W., Elementary Applied Topology (CreateS- pace, Scotts Valley, CA, 2016)

    2016

  59. [67]

    Sun, H., Radicchi, F., Kurths, J., Bianconi, G., The dy- namic nature of percolation on networks with triadic in- teractions. Nat. Commun. 14, 1308 (2023)

    2023

  60. [68]

    We explore how to streamline this process by leveraging the statisti- cal properties of the GBS output distribution

    and even for quantum computers [22]. We explore how to streamline this process by leveraging the statisti- cal properties of the GBS output distribution. Because clique percolation fundamentally captures the essence of network connectivity, analysing GBS outcomes with an entro...

  61. [69]

    & Vicsek, T

    Palla, G., Der´ enyi, I., Farkas, I. & Vicsek, T. Uncover- ing the overlapping community structure of complex net- works in nature and society. Nature 435, 814-818 (2005)

    2005

  62. [70]

    & Parigi, V

    Nokkala, J., Arzani, F., Galve, F., Zambrini, R., Man- iscalco, S., Piilo, J., Treps, N. & Parigi, V. Reconfig- urable optical implementation of quantum complex net- works.New Journal of Physics20, 053024 (2018)

    2018

  63. [71]

    Topological percolation on hyperbolic sim- plicial complexes

    Bianconi, G. Topological percolation on hyperbolic sim- plicial complexes. Phys. Rev. E98, 052308 (2018)

    2018

  64. [72]

    & Amico, E

    Santoro, A., Battiston, F., Lucas, M., Petri, G. & Amico, E. Higher-order connectomics of human brain function reveals local topological signatures of task decoding, in- dividual identification, and behavior. Nat. Commun. 15, 10244 (2024)

    2024

  65. [73]

    Hasan, M. Z. & Kane, C. L. Colloquium: Topological insulators. Rev. Mod. Phys.82, 3045 (2010)

    2010

  66. [74]

    M., Kim, F., Penumadu, D

    Thakur, M. M., Kim, F., Penumadu, D. & Herring, A. Pore space and fluid phase characterization in round and angular partially saturated sands using radiation- based tomography and persistent homology. Transport in Porous Media137, 131-155 (2021)

    2021

  67. [75]

    & Inoue, S., Non-Gaussian operation based on photon subtraction using a photon-number-resolving detector at a telecommunications wavelength

    Namekata, N., Takahashi, Y., Fujii, G., Fukuda, D., Kurimura, S. & Inoue, S., Non-Gaussian operation based on photon subtraction using a photon-number-resolving detector at a telecommunications wavelength. Nat. Pho- ton. 4, 655-660 (2010)

    2010

  68. [76]

    Konno, S. et. al. Logical states for fault-tolerant quantum computation with propagating light, Science 383, 289- 293 (2024)

    2024

  69. [77]

    Putterman, H. et. al. Hardware-efficient quantum error correction via concatenated bosonic qubits. Nature 638, 927-934 (2025)

    2025

  70. [78]

    Mill´ an, A. P. et. al. Topology shapes dynamics of higher- order networks. Nat. Phys. 21, 353-361 (2025). Acknowledgments We thank Nicol´ as Quesada and Changhun Oh for the helpful discussions and feedback on the manuscript. This work was supported by UK Research and In- nova...

    2025

Pith tools

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