arxiv: 2605.00772 · v1 · submitted 2026-05-01 · 🪐 quant-ph · cond-mat.stat-mech

Recognition: unknown

Entanglement capacity of complex networks from quantum walks

Pravy Prerana , Sascha Wald

Authors on Pith no claims yet

Pith reviewed 2026-05-09 19:23 UTC · model grok-4.3

classification 🪐 quant-ph cond-mat.stat-mech

keywords quantum walksentanglementcomplex networksgraph matchingssource-target bipartitionquantum transportrandom graphs

0 comments

The pith

Network connectivity imposes an upper bound on entanglement from quantum walks, set by graph matchings.

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

The paper develops source-target entanglement as a measure for quantum walks on arbitrary networks by giving every node both source and target roles. This measure is shown to be strictly limited by the network's connectivity, with the size of a maximum matching determining the bound. The work further demonstrates that in random graphs, adding edges and improving connectivity reduces the maximum entanglement that can be achieved.

Core claim

For discrete-time quantum walks on general networks, the source-target entanglement defined by a bipartition that assigns each node dual source and target roles is upper-bounded by the network connectivity, with graph matchings as the underlying structure governing entanglement generation, and increasing connectivity in random graphs decreases the attainable entanglement.

What carries the argument

Source-target entanglement, obtained by bipartitioning each node into source and target roles to generalize the coin-walker entanglement concept to irregular graphs.

If this is right

Network connectivity imposes an upper bound on the attainable entanglement.
Graph matchings are the structures that control entanglement generation.
In random graphs, greater connectivity reduces the maximum entanglement.

Where Pith is reading between the lines

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

Quantum walk protocols on networks may achieve higher entanglement with sparser rather than denser topologies.
Combinatorial tools from graph theory can be used to predict and bound entanglement capacity in walk-based quantum systems.

Load-bearing premise

Assigning each node both source and target roles yields a physically meaningful entanglement measure that extends the coin-walker notion to irregular networks.

What would settle it

A concrete computation on an irregular network in which the measured source-target entanglement exceeds the value set by the size of its maximum matching would refute the claimed upper bound.

Figures

Figures reproduced from arXiv: 2605.00772 by Pravy Prerana, Sascha Wald.

**Figure 1.** Figure 1: FIG. 1 view at source ↗

**Figure 2.** Figure 2: FIG. 2 view at source ↗

**Figure 3.** Figure 3: (b)) although BA networks will remain structurally heterogeneous. To relate the dynamical results to structural properties of the underlying random graphs, we time-average the plateau regime for each realization and compute the average clustering coefficient of the corresponding network. The clustering coefficient quantifies the density of triangles, i.e. the fraction of neighbor pairs of a node that are… view at source ↗

read the original abstract

Discrete-time quantum walks provide a natural framework for quantum transport on complex networks. On regular structures, coin-walker entanglement has been widely used to characterize quantum transport and to support quantum algorithmic protocols. However, this notion relies on a fixed Hilbert space factorization separating coin and position and is therefore not directly applicable to more complex, irregular structures. Here we introduce an entanglement measure for general networks based on a bipartition that assigns each node two roles, acting as both a source and a target. The resulting bipartition defines the source-target entanglement, a measure for general networks, motivated by coin-walker entanglement. We show that the connectivity of the network imposes an upper bound on this entanglement and identify graph matchings as the underlying structure governing entanglement generation. We further illustrate that in random graphs improving graph connectivity reduces the attainable entanglement, establishing a structure-dependent constraint on quantum correlations.

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.

Desk Editor's Note private letter to a colleague

The paper introduces a source-target entanglement for quantum walks on irregular networks and links it to graph matchings plus a connectivity upper bound, but the bipartition's Hilbert-space validity is the open question.

read the letter

The main takeaway is that the authors replace the usual coin-position split with a source-target bipartition on each node to define an entanglement measure that works on general graphs. They then argue that network connectivity caps this entanglement and that perfect matchings in the graph control how much can be generated. The random-graph example shows that adding edges tends to reduce the attainable value, which is a concrete and somewhat counter-intuitive result.

Referee Report

2 major / 2 minor

Summary. The manuscript introduces a source-target entanglement measure for discrete-time quantum walks on general complex networks by assigning each node dual source and target roles to induce a bipartition of the Hilbert space. It claims that network connectivity imposes an upper bound on this entanglement, identifies graph matchings as the combinatorial structure governing entanglement generation, and illustrates via random graphs that higher connectivity reduces the maximum attainable entanglement.

Significance. If the bipartition is rigorously defined and the bounds hold, the work provides a structure-dependent constraint on quantum correlations generated by walks on irregular topologies, extending coin-walker entanglement concepts and linking them to graph-theoretic objects such as matchings. This could inform quantum transport protocols and algorithmic applications on complex networks.

major comments (2)

[§3] §3 (Definition of source-target bipartition): The construction duplicates nodes into source and target copies to define the bipartition, but on irregular graphs the coin spaces have node-dependent dimensions. An explicit isometric embedding or identification map must be provided to ensure the total Hilbert space factors as a tensor product H_source ⊗ H_target on which the walk operator acts unitarily; without this, the partial trace does not necessarily yield a valid entanglement monotone, undermining the claimed upper bounds and matching interpretation.
[§4] §4 (Upper bound theorem): The statement that connectivity imposes an upper bound on source-target entanglement is load-bearing for the central claim, yet the proof sketch relies on the matching structure without specifying whether the bound holds for all initial states, for time-averaged entanglement, or only for specific walks. A counter-example on a small irregular graph (e.g., a path with an added edge) should be checked to confirm the bound is not an artifact of the factorization.

minor comments (2)

[Figure 2] Figure 2 caption: Clarify whether the plotted quantity is the maximum entanglement over time or the time-averaged value, and specify the ensemble (Erdős–Rényi or otherwise) and number of realizations used for the random-graph data.
[§3] Notation: The symbol for the reduced density matrix after tracing out the target subspace should be introduced once and used consistently; currently the partial-trace operation is described in prose without an equation label.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive suggestions. We address each major comment below and have revised the manuscript to strengthen the rigor of the definitions and proofs.

read point-by-point responses

Referee: [§3] §3 (Definition of source-target bipartition): The construction duplicates nodes into source and target copies to define the bipartition, but on irregular graphs the coin spaces have node-dependent dimensions. An explicit isometric embedding or identification map must be provided to ensure the total Hilbert space factors as a tensor product H_source ⊗ H_target on which the walk operator acts unitarily; without this, the partial trace does not necessarily yield a valid entanglement monotone, undermining the claimed upper bounds and matching interpretation.

Authors: We agree that an explicit isometric embedding is required to rigorously define the tensor-product factorization on irregular graphs. In the revised manuscript we introduce a canonical embedding: each local coin space C_v of dimension deg(v) is isometrically embedded into a fixed auxiliary space of dimension equal to the maximum degree by zero-padding. This yields a uniform identification map under which the total Hilbert space factors as H_source ⊗ H_target, the duplicated-node construction becomes a well-defined bipartition, and the walk operator remains unitary. The partial trace is therefore a valid quantum operation, and the source-target entanglement is a proper monotone. We have expanded §3 with the full mathematical construction, a proof that the embedding preserves the walk dynamics, and a worked example on a small irregular graph. revision: yes
Referee: [§4] §4 (Upper bound theorem): The statement that connectivity imposes an upper bound on source-target entanglement is load-bearing for the central claim, yet the proof sketch relies on the matching structure without specifying whether the bound holds for all initial states, for time-averaged entanglement, or only for specific walks. A counter-example on a small irregular graph (e.g., a path with an added edge) should be checked to confirm the bound is not an artifact of the factorization.

Authors: The upper bound applies to the entanglement capacity, i.e., the supremum of the source-target entanglement over all initial states and all evolution times. The proof shows that no quantum walk trajectory can generate more entanglement than the size of a maximum matching, because each matching edge is the only combinatorial object that can create the necessary source-target correlations; the bound is therefore independent of the particular initial state and holds at every time step (hence also for any time average). We have clarified this scope explicitly in the revised §4. In addition, we have computed the full entanglement dynamics on the suggested counter-example (a path of four vertices with one added chord) and verified that the observed maximum never exceeds the matching number. This numerical check is now included as a new example in the manuscript. revision: yes

Circularity Check

0 steps flagged

No circularity in the derivation chain

full rationale

The paper introduces a source-target bipartition to define an entanglement measure for irregular networks, extending the coin-walker concept, then derives an upper bound on entanglement from network connectivity and identifies graph matchings as the governing structure via analysis of the quantum walk operator. These results follow directly from the mathematical properties of the defined Hilbert space and walk dynamics on the graph, without any reduction of the claimed bounds or predictions to fitted parameters, self-definitional equivalences, or load-bearing self-citations. The central claims remain independent of the inputs and are presented as consequences of network topology rather than tautological restatements.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract provides no explicit free parameters, axioms, or invented entities; the work is presented as a conceptual extension using standard quantum mechanics and graph theory.

pith-pipeline@v0.9.0 · 5440 in / 1138 out tokens · 29341 ms · 2026-05-09T19:23:14.017719+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

71 extracted references · 8 canonical work pages

[1]

Aharonov, L

Y. Aharonov, L. Davidovich, and N. Zagury, Phys. Rev. A48, 1687 (1993)

1993
[2]

Kempe, Contemporary Physics44, 307 (2003)

J. Kempe, Contemporary Physics44, 307 (2003)

2003
[3]

Aharonov, A

D. Aharonov, A. Ambainis, J. Kempe, and U. Vazirani, inProceedings of the thirty-third annual ACM symposium on Theory of computing, STOC01 (ACM, 2001) p. 50–59

2001
[4]

S. E. Venegas-Andraca, Quantum Information Process- ing11, 1015–1106 (2012)

2012
[5]

F. Xia, J. Liu, H. Nie, Y. Fu, L. Wan, and X. Kong, IEEE Transactions on Emerging Topics in Computational In- telligence4, 95 (2020)

2020
[6]

Kadian, S

K. Kadian, S. Garhwal, and A. Kumar, Computer Sci- ence Review41, 100419 (2021)

2021
[7]

Karski, L

M. Karski, L. F¨ orster, J.-M. Choi, A. Steffen, W. Alt, D. Meschede, and A. Widera, Science325, 174–177 (2009)

2009
[8]

L. H. Yao and S. Wald, Phys. Rev. E108, 024139 (2023)

2023
[9]

A. M. Childs, Phys. Rev. Lett.102, 180501 (2009)

2009
[10]

N. B. Lovett, S. Cooper, M. Everitt, M. Trevers, and V. Kendon, Phys. Rev. A81, 042330 (2010)

2010
[11]

Yamagishi, N

M. Yamagishi, N. Hatano, and H. Obuse, Scientific Re- ports14, 10.1038/s41598-024-78986-z (2024)

work page doi:10.1038/s41598-024-78986-z 2024
[12]

Gipouloux, M

J. Gipouloux, M. Brunelli, L. Cugliandolo, R. Fazio, and M. Schir` o, Active quantum particles from engineered dis- sipation (2026), arXiv:2603.19094 [quant-ph]

work page arXiv 2026
[13]

Khasseh, S

R. Khasseh, S. Wald, R. Moessner, C. A. Weber, and M. Heyl, Phys. Rev. Lett.135, 248302 (2025)

2025
[14]

Nadolny, C

T. Nadolny, C. Bruder, and M. Brunelli, Phys. Rev. X 15, 011010 (2025)

2025
[15]

A. P. Antonov, Y. Zheng, B. Liebchen, and H. L¨ owen, Phys. Rev. Res.7, 033008 (2025)

2025
[16]

A. P. Antonov, S. Lee, B. Liebchen, H. L¨ owen, J. Melles, G. Morigi, Y. Tuchkov, and M. te Vrugt, Modeling dissipation in quantum active matter (2026), arXiv:2511.21502 [quant-ph]

work page arXiv 2026
[17]

Vieira, E

R. Vieira, E. P. M. Amorim, and G. Rigolin, Phys. Rev. Lett.111, 180503 (2013)

2013
[18]

Vieira, E

R. Vieira, E. P. M. Amorim, and G. Rigolin, Phys. Rev. A89, 042307 (2014)

2014
[19]

S´ anchez-Burillo, J

E. S´ anchez-Burillo, J. Duch, J. G´ omez-Garde˜ nes, and D. Zueco, Scientific Reports2, 10.1038/srep00605 (2012)

work page doi:10.1038/srep00605 2012
[20]

Shenvi, J

N. Shenvi, J. Kempe, and K. B. Whaley, Phys. Rev. A 67, 052307 (2003)

2003
[21]

A. M. Childs and J. Goldstone, Phys. Rev. A70, 022314 (2004)

2004
[22]

Tulsi, Phys

A. Tulsi, Phys. Rev. A78, 012310 (2008)

2008
[23]

L. Xu, M. Li, and X. Sun, Phys. Rev. A112, 062422 (2025)

2025
[24]

Innocenti, H

L. Innocenti, H. Majury, T. Giordani, N. Spagnolo, F. Sciarrino, M. Paternostro, and A. Ferraro, Phys. Rev. A96, 062326 (2017)

2017
[25]

Giordani, E

T. Giordani, E. Polino, S. Emiliani, A. Suprano, L. Inno- centi, H. Majury, L. Marrucci, M. Paternostro, A. Fer- raro, N. Spagnolo, and F. Sciarrino, Phys. Rev. Lett.122, 020503 (2019)

2019
[26]

Srikara and C

S. Srikara and C. M. Chandrashekar, Quantum Informa- tion Processing19, 10.1007/s11128-020-02793-4 (2020)

work page doi:10.1007/s11128-020-02793-4 2020
[27]

Giordani, L

T. Giordani, L. Innocenti, A. Suprano, E. Polino, M. Pa- ternostro, N. Spagnolo, F. Sciarrino, and A. Ferraro, New Journal of Physics23, 023012 (2021)

2021
[28]

S. S. Panda, P. A. A. Yasir, and C. M. Chandrashekar, Phys. Rev. A107, 022611 (2023)

2023
[29]

Chawla, A

P. Chawla, A. Ajith, and C. M. Chandrashekar, Physica Scripta98, 105113 (2023)

2023
[30]

S. Das, A. V. N. S. Meghnath, S. Rajiuddin, and P. K. Panigrahi, Quantum Information Processing24, 271 (2025)

2025
[31]

Kitagawa, M

T. Kitagawa, M. A. Broome, A. Fedrizzi, M. S. Rud- ner, E. Berg, I. Kassal, A. Aspuru-Guzik, E. Dem- ler, and A. G. White, Nature Communications3, 6 10.1038/ncomms1872 (2012)

work page doi:10.1038/ncomms1872 2012
[32]

K. Wang, X. Qiu, L. Xiao, X. Zhan, Z. Bian, W. Yi, and P. Xue, Phys. Rev. Lett.122, 020501 (2019)

2019
[33]

D. K. Panda and C. Benjamin, Phys. Rev. B , (2026)

2026
[34]

D. K. Panda and C. Benjamin, Fractional Topological phases, flat bands, and robust edge states on finite cyclic graphs via single-coin split-step quantum walks (2026), arXiv:2603.07701 [cond-mat.str-el]

work page arXiv 2026
[35]

Mukherjee, R

A. Mukherjee, R. A. R¨ omer, and A. Chakrabarti, Phys. Rev. B100, 161108 (2019)

2019
[36]

Manouchehri and J

K. Manouchehri and J. Wang,Physical Implementation of Quantum Walks(Springer Berlin Heidelberg, 2014)

2014
[37]

P. L. Knight, E. Rold´ an, and J. E. Sipe, Phys. Rev. A 68, 020301 (2003)

2003
[38]

P. L. Knight, E. Rold´ an, and J. Sipe, Optics Communi- cations227, 147–157 (2003)

2003
[39]

M. C. Ba˜ nuls, C. Navarrete, A. P´ erez, E. Rold´ an, and J. C. Soriano, Phys. Rev. A73, 062304 (2006)

2006
[40]

Eckert, J

K. Eckert, J. Mompart, G. Birkl, and M. Lewenstein, Phys. Rev. A72, 012327 (2005)

2005
[41]

C. M. Chandrashekar, Phys. Rev. A74, 032307 (2006)

2006
[42]

Koˇ s´ ık and V

J. Koˇ s´ ık and V. Buˇ zek, Phys. Rev. A71, 012306 (2005)

2005
[43]

A. P. Hines and P. C. E. Stamp, Phys. Rev. A75, 062321 (2007)

2007
[44]

B. L. Douglas and J. B. Wang, Phys. Rev. A79, 052335 (2009)

2009
[45]

Loke and J

T. Loke and J. Wang, Computer Physics Communica- tions182, 2285–2294 (2011)

2011
[46]

Qiang, Y

X. Qiang, Y. Wang, S. Xue, R. Ge, L. Chen, Y. Liu, A. Huang, X. Fu, P. Xu, T. Yi, F. Xu, M. Deng, J. B. Wang, J. D. A. Meinecke, J. C. F. Matthews, X. Cai, X. Yang, and J. Wu, Science Advances7, eabb8375 (2021)

2021
[47]

B¨ ottcher and M

L. B¨ ottcher and M. A. Porter, SIAM Journal on Applied Mathematics81, 2704–2724 (2021)

2021
[48]

Wald and L

S. Wald and L. B¨ ottcher, Phys. Rev. E103, 012122 (2021)

2021
[49]

Acasiete, F

F. Acasiete, F. P. Agostini, J. K. Moqadam, and R. Por- tugal, Quantum Information Processing19, 426 (2020)

2020
[50]

Xu, X.-W

X.-Y. Xu, X.-W. Wang, D.-Y. Chen, C. M. Smith, and X.-M. Jin, Nature Photonics15, 703 (2021)

2021
[51]

Krovi and T

H. Krovi and T. A. Brun, Phys. Rev. A74, 042334 (2006)

2006
[52]

Krovi and T

H. Krovi and T. A. Brun, Phys. Rev. A75, 062332 (2007)

2007
[53]

Portugal,Quantum Walks and Search Algorithms (Springer International Publishing, 2018)

R. Portugal,Quantum Walks and Search Algorithms (Springer International Publishing, 2018)

2018
[54]

T. G. Wong, Quantum Information Processing16, 215 (2017)

2017
[55]

Portugal, Quantum Information Processing15, 1387 (2016)

R. Portugal, Quantum Information Processing15, 1387 (2016)

2016
[56]

Portugal and E

R. Portugal and E. Segawa, Interdisciplinary Information Sciences23, 119 (2017)

2017
[57]

M. Xu, Z. Liu, H. Chen, and S. Zheng, Quantum Infor- mation Processing20, 51 (2021)

2021
[58]

M. A. Nielsen and I. L. Chuang,Quantum Computation and Quantum Information, 10th ed. (Cambridge Univer- sity Press, Cambridge, UK, 2010)

2010
[59]

Sperling and W

J. Sperling and W. Vogel, Physica Scripta83, 045002 (2011)

2011
[60]

Quantum Walk on the Line

A. Nayak and A. Vishwanath, Quantum walk on the line (2000), arXiv:quant-ph/0010117 [quant-ph]

work page Pith review arXiv 2000
[61]

See Supplemental Material for details
[62]

Albert and A.-L

R. Albert and A.-L. Barab´ asi, Rev. Mod. Phys.74, 47 (2002)

2002
[63]

R. M. Karp and M. Sipser, in22nd Annual Symposium on Foundations of Computer Science (sfcs 1981)(IEEE,

1981
[64]

Zdeborov´ a and M

L. Zdeborov´ a and M. M´ ezard, Journal of Statistical Me- chanics: Theory and Experiment2006, P05003 (2006)

2006
[65]

Jahnke, J

L. Jahnke, J. W. Kantelhardt, R. Berkovits, and S. Havlin, Phys. Rev. Lett.101, 175702 (2008)

2008
[66]

Chakraborty, L

S. Chakraborty, L. Novo, A. Ambainis, and Y. Omar, Phys. Rev. Lett.116, 100501 (2016)

2016
[67]

Yin and E

R. Yin and E. Barkai, Phys. Rev. Lett.130, 050802 (2023)

2023
[68]

Prerana and S

P. Prerana and S. Wald, Zenodo (2026). 1 SUPPLEMENTAL MATERIAL Entanglement capacity of complex networks from quantum walks Pravy Prerana1 and Sascha Wald1 1Centre for Fluid and Complex Systems, Coventry University, Coventry, CV1 2TT, United Kingdom COIN ASSIGNMENT IN THE HADAMARD WALK FIG. S1.Coin-walker representation for the Hadamard walk.(a) Schematic...

2026
[69]

This setup is illustrated in Fig. S1(a). Clearly the line is two-regular and has edge chromatic number 2. However, the above prescription for the Hadamard walk assigns a particular edge two distinct coin states depending on whether it is traversed leftward or rightward. In this representation, the coin space is introduced independently of the coloring of ...
[70]

In the standard coin-walker notation this state can be written as|ϕ⟩= (|0↑⟩+|2↓⟩)/ √ 2, admitting maximum coin-walker entanglement

In this manner, the DTQW state is well-defined with components that reside on sites 0 and 2 and intend to move to site 1. In the standard coin-walker notation this state can be written as|ϕ⟩= (|0↑⟩+|2↓⟩)/ √ 2, admitting maximum coin-walker entanglement. If we however consider the coin description given in Fig. S1(b), i.e.|ϕ CW⟩= (|0↑⟩+|2↑⟩)/ √ 2 we see th...
[71]

We see that both quantities behave quite differently: Although both measures show a sharp initial rise with subsequent leveling off, not only are the scales different, also the coin-walker entanglement quickly reaches its stationary value and does not indicate any walker dynamics any more, whereas the source-target entanglement still clearly evidences the...