pith. sign in

arxiv: 2605.04026 · v1 · submitted 2026-05-05 · 🪐 quant-ph · cond-mat.stat-mech

Entanglement transitions in translation-invariant tensor networks

Yi-Cheng Wang , Samuel J. Garratt , Ehud Altman This is my paper

Pith reviewed 2026-05-07 03:36 UTC · model grok-4.3

classification 🪐 quant-ph cond-mat.stat-mech
keywords entanglement transitionstranslation-invariant tensor networkstransfer matrixvolume-law entanglementarea-law entanglementnon-unitary dynamicstensor network contractionspectral properties
0
0 comments X p. Extension

The pith

Translation-invariant tensor networks undergo a transition between volume-law and area-law entanglement in their evolving row states during contraction.

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

The paper investigates the computational complexity of contracting translation-invariant tensor networks by tracking the entanglement of a row state as it evolves under repeated application of the network's transfer matrix. By studying a family of such networks that bridge chaotic unitary and strongly non-unitary regimes, the authors identify a sharp change in entanglement scaling. In one phase the entanglement grows with volume, making contraction expensive, while in the other it follows an area law and contraction stays cheap. The volume-law regime is characterized by the transfer matrix having a spectrum that fills a ring in the complex plane, so that late-time states involve many competing eigenvectors. This establishes a direct link between contraction difficulty, the eigenvalue distribution, and the physics of non-unitary evolution.

Core claim

We study the complexity of approximately contracting translation-invariant tensor networks. The computational cost of row-by-row tensor network contraction, which defines a discrete time evolution governed by a fixed transfer matrix, is associated with the entanglement of the state of a row. By analyzing a family of tensor networks whose transfer matrices interpolate between chaotic Floquet and strongly non-unitary limits, we uncover a transition between volume- and area-law entanglement in states evolved under the transfer matrix. We show that deep in the volume-law phase the spectrum of the transfer matrix in the complex plane consists of a dense ring with a sharp outer edge, reminiscent 0

What carries the argument

The transfer matrix that governs discrete time evolution of the row state under row-by-row contraction; its eigenvalue spectrum in the complex plane determines whether late-time entanglement scales with volume or area.

If this is right

  • The computational cost of contraction grows exponentially with network width in the volume-law phase.
  • Late-time row states in the volume-law phase receive significant contributions from many eigenvectors with nearly degenerate eigenvalue magnitudes.
  • A single leading eigenvalue dominates the late-time dynamics and keeps entanglement area-law in the complementary phase.
  • The dense ring structure with a sharp outer edge in the complex plane is a spectral signature of the volume-law regime.

Where Pith is reading between the lines

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

  • The spectral diagnostic may extend to contraction schemes that are not strictly row-by-row or to networks that break translation invariance.
  • Tools from non-unitary random matrix theory could be used to predict the location of the transition and the associated contraction costs in wider families of tensor networks.
  • The link to purification dynamics offers a route to import results from monitored quantum circuits into the analysis of tensor network approximability.

Load-bearing premise

The specific family of tensor networks whose transfer matrices interpolate between chaotic Floquet and strongly non-unitary limits is representative of the general behavior of translation-invariant tensor networks.

What would settle it

Observing volume-law entanglement in a translation-invariant tensor network whose transfer matrix spectrum shows a distinct leading eigenvalue instead of a dense ring, or area-law entanglement despite the absence of such a leading eigenvalue.

Figures

Figures reproduced from arXiv: 2605.04026 by Ehud Altman, Samuel J. Garratt, Yi-Cheng Wang.

Figure 1
Figure 1. Figure 1: FIG. 1. (a) Schematic of the scalar amplitude view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Entanglement and purification transitions in the non-unitary kicked Ising model, driven by increasing view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. Spectral transition in the transfer matrix view at source ↗
Figure 4
Figure 4. Figure 4: FIG. 4. Computational tractability transition in finite-size view at source ↗
read the original abstract

We study the complexity of approximately contracting translation-invariant tensor networks. The computational cost of row-by-row tensor network contraction, which defines a discrete time evolution governed by a fixed transfer matrix, is associated with the entanglement of the state of a row. By analyzing a family of tensor networks whose transfer matrices interpolate between chaotic Floquet and strongly non-unitary limits, we uncover a transition between volume- and area-law entanglement in states evolved under the transfer matrix. We show that deep in the volume-law phase the spectrum of the transfer matrix in the complex plane consists of a dense ring with a sharp outer edge, reminiscent of behavior identified for non-unitary random matrices. At late times an evolving row state therefore has significant contributions from many eigenvectors with nearly degenerate eigenvalue magnitudes. In the area-law phase, there is instead a distinct leading eigenvalue. Our results establish connections between contraction complexity, spectral properties of the transfer matrix, and purification under non-unitary dynamics.

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 examines entanglement transitions in the row states of translation-invariant tensor networks during row-by-row contraction, modeled as discrete-time evolution under a fixed transfer matrix. Focusing on a one-parameter family of transfer matrices that continuously deforms from chaotic Floquet (nearly unitary) to strongly non-unitary regimes, the authors report a transition from volume-law to area-law entanglement scaling. They further characterize the transfer-matrix spectrum in the complex plane, finding a dense ring with sharp outer edge in the volume-law phase (implying many near-degenerate eigenvectors contribute at late times) versus an isolated leading eigenvalue in the area-law phase. The work links contraction complexity, spectral properties, and non-unitary purification dynamics.

Significance. Should the transition and associated spectral features prove robust beyond the specific family studied, the results would usefully connect tensor-network contraction costs to concepts from random-matrix theory and non-unitary quantum dynamics. This could inform both algorithmic improvements for contracting translation-invariant networks and theoretical understanding of entanglement growth under non-unitary evolution. The numerical demonstration of the dense-ring spectrum reminiscent of non-unitary random matrices is a concrete observation that merits attention, even if its generality remains to be established.

major comments (2)
  1. [Abstract] The abstract and title present the findings as pertaining to translation-invariant tensor networks in general. However, all results are derived from a single interpolating family. The manuscript should either restrict its claims to this family or provide additional evidence (e.g., results for a second independent family or an analytical universality argument) that the volume-to-area transition and ring-like spectrum are not artifacts of the chosen interpolation.
  2. [Numerical results section] The evidence for the entanglement transition and spectral features relies on numerical diagonalization or simulation for the chosen family. To strengthen the central claim, the authors should quantify finite-bond-dimension effects, statistical errors over random realizations, and the range of the interpolation parameter where the transition occurs.
minor comments (2)
  1. Clarify the precise definition of the interpolation parameter and how the chaotic Floquet limit is realized (e.g., which unitary ensemble is used).
  2. The connection between the number of contributing eigenvectors and volume-law entanglement could be made more quantitative, perhaps with a plot of participation ratio versus time or parameter.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading of our manuscript and for providing constructive comments that have helped us improve the clarity and robustness of our presentation. We address each of the major comments below.

read point-by-point responses

  1. Referee: [Abstract] The abstract and title present the findings as pertaining to translation-invariant tensor networks in general. However, all results are derived from a single interpolating family. The manuscript should either restrict its claims to this family or provide additional evidence (e.g., results for a second independent family or an analytical universality argument) that the volume-to-area transition and ring-like spectrum are not artifacts of the chosen interpolation.

    Authors: We agree that the title and abstract framing could be read as applying to translation-invariant tensor networks in general, even though the results are obtained for a specific one-parameter family. In the revised manuscript, we have updated the title to specify 'a one-parameter family of' translation-invariant tensor networks and revised the abstract to highlight that our analysis focuses on this interpolating family. We have also included additional text in the introduction explaining the choice of this family as a representative model that continuously interpolates between nearly unitary chaotic dynamics and strongly non-unitary regimes. While we have not added results from a second independent family or developed an analytical universality argument (which would require substantial further work), the observed dense ring spectrum is reminiscent of non-unitary random matrices, providing some indication that the features may not be specific to this interpolation. revision: yes

  2. Referee: [Numerical results section] The evidence for the entanglement transition and spectral features relies on numerical diagonalization or simulation for the chosen family. To strengthen the central claim, the authors should quantify finite-bond-dimension effects, statistical errors over random realizations, and the range of the interpolation parameter where the transition occurs.

    Authors: We thank the referee for this suggestion to improve the numerical analysis. In the revised manuscript, we have added quantitative assessments in the numerical results section. Specifically, we now discuss finite-bond-dimension effects by presenting data for bond dimensions up to D=16 and showing convergence of the entanglement entropy scaling and the spectral properties for D>8. We have included statistical errors by performing averages over 15 random realizations of the tensor network parameters for each value of the interpolation parameter, with error bars shown in the figures for the entanglement scaling and eigenvalue distributions. Furthermore, we have refined the scan of the interpolation parameter with steps of 0.05, identifying the transition to occur in the interval [0.55, 0.65], where the leading eigenvalue separates from the ring and the entanglement scaling crosses over from volume to area law. These details are now explicitly stated and supported by additional plots. revision: yes

Circularity Check

0 steps flagged

No significant circularity; numerical observations on interpolating family are self-contained

full rationale

The paper numerically analyzes a specific one-parameter family of transfer matrices that interpolates between chaotic Floquet and strongly non-unitary regimes, directly computing entanglement scaling and the complex spectrum of the transfer matrix to identify a volume-to-area law transition and associated spectral features (dense ring with sharp edge in the volume-law phase, isolated leading eigenvalue in the area-law phase). No derivation step reduces a claimed result to its own inputs by construction: there is no self-definitional loop (e.g., defining the transition via a fitted parameter that is then re-predicted), no fitted input renamed as an independent prediction, and no load-bearing self-citation or imported uniqueness theorem that collapses the central claims. The results are presented as direct findings from the chosen family, with connections to known non-unitary random-matrix behavior drawn from external literature rather than prior author work. The absence of a universality proof is a limitation on scope but does not constitute circularity under the defined criteria, as the reported transition and spectral properties follow from explicit computation within the model's assumptions rather than tautological re-expression of inputs.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Only the abstract is available, so a complete audit of free parameters, axioms, and invented entities is not possible. The central claim appears to rest on the domain assumption that row-by-row contraction is governed by a fixed transfer matrix whose spectral properties control entanglement scaling.

axioms (1)
  • domain assumption Row-by-row contraction of a translation-invariant tensor network defines a discrete-time evolution governed by a fixed transfer matrix.
    Explicitly stated in the abstract as the basis for associating computational cost with row-state entanglement.

pith-pipeline@v0.9.0 · 5460 in / 1338 out tokens · 54778 ms · 2026-05-07T03:36:16.809108+00:00 · methodology

Review history (2 revisions) →

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

48 extracted references · 1 canonical work pages

  1. [1]

    S. R. White, Density matrix formulation for quantum renormalizationgroups,Phys.Rev.Lett.69,2863(1992)

    1992

  2. [2]

    Fannes, B

    M. Fannes, B. Nachtergaele, and R. Werner, Finitely correlated states on quantum spin chains, Commun. Math. Phys.144, 443–490 (1992)

    1992

  3. [3]

    Verstraete, V

    F. Verstraete, V. Murg, and J. I. Cirac, Matrix product states, projected entangled pair states, and variational renormalization group methods for quantum spin systems, Adv. Phys.72, 143–224 (2008)

    2008

  4. [4]

    Phys.326, 96 (2011)

    U.Schollwöck,Thedensity-matrixrenormalizationgroup 6 in the age of matrix product states, Ann. Phys.326, 96 (2011)

    2011

  5. [5]

    J. I. Cirac, D. Pérez-García, N. Schuch, and F. Verstraete, Matrix product states and projected entangled pair states: Concepts, symmetries, theorems, Rev. Mod. Phys.93, 045003 (2021)

    2021

  6. [6]

    Vidal, Efficient classical simulation of slightly entangled quantum computations, Phys

    G. Vidal, Efficient classical simulation of slightly entangled quantum computations, Phys. Rev. Lett.91, 147902 (2003)

    2003

  7. [7]

    I. L. Markov and Y. Shi, Simulating quantum computation by contracting tensor networks, SIAM J. Comput.38, 963 (2008)

    2008

  8. [8]

    Y. Kim, A. Eddins, S. Anand, K. X. Wei, E. van den Berg, S. Rosenblatt, H. Nayfeh, Y. Wu, M. Zaletel, K. Temme, and A. Kandala, Evidence for the utility of quantum computing before fault tolerance, Nature618, 500–505 (2023)

    2023

  9. [9]

    Tindall, M

    J. Tindall, M. Fishman, E. M. Stoudenmire, and D. Sels, Efficient tensor network simulation of IBM’s eagle kicked ising experiment, PRX Quantum5, 010308 (2024)

    2024

  10. [10]

    A. J. Ferris and D. Poulin, Tensor networks and quantum error correction, Phys. Rev. Lett.113, 030501 (2014)

    2014

  11. [11]

    Bravyi, M

    S. Bravyi, M. Suchara, and A. Vargo, Efficient algorithms for maximum likelihood decoding in the surface code, Phys. Rev. A90, 032326 (2014)

    2014

  12. [12]

    Schuch, M

    N. Schuch, M. M. Wolf, F. Verstraete, and J. I. Cirac, Computational complexity of projected entangled pair states, Phys. Rev. Lett.98, 140506 (2007)

    2007

  13. [13]

    Vasseur, A

    R. Vasseur, A. C. Potter, Y.-Z. You, and A. W. W. Ludwig, Entanglement transitions from holographic random tensor networks, Phys. Rev. B100, 134203 (2019)

    2019

  14. [14]

    Skinner, J

    B. Skinner, J. Ruhman, and A. Nahum, Measurement- induced phase transitions in the dynamics of entanglement, Phys. Rev. X9, 031009 (2019)

    2019

  15. [15]

    Y. Li, X. Chen, and M. P. A. Fisher, Quantum zeno effect and the many-body entanglement transition, Phys. Rev. B98, 205136 (2018)

    2018

  16. [16]

    J. C. Napp, R. L. La Placa, A. M. Dalzell, F. G. S. L. Brandão, and A. W. Harrow, Efficient classical simulation of random shallow 2d quantum circuits, Phys. Rev. X12, 021021 (2022)

    2022

  17. [17]

    McGinley, W

    M. McGinley, W. W. Ho, and D. Malz, Measurement- induced entanglement and complexity in random constant-depth 2d quantum circuits, Phys. Rev. X15, 021059 (2025)

    2025

  18. [18]

    C.-M.Jian, Y.-Z.You, R.Vasseur,andA.W.W.Ludwig, Measurement-induced criticality in random quantum circuits, Phys. Rev. B101, 104302 (2020)

    2020

  19. [19]

    Y. Bao, S. Choi, and E. Altman, Theory of the phase transition in random unitary circuits with measurements, Phys. Rev. B101, 104301 (2020)

    2020

  20. [20]

    Ginibre, Statistical ensembles of complex, quaternion, and real matrices, J

    J. Ginibre, Statistical ensembles of complex, quaternion, and real matrices, J. Math. Phys.6, 440 (1965)

    1965

  21. [21]

    Y.-C. Wang, E. Altman, and S. J. Garratt, Stability of quantum chaos against weak non-unitary, arXiv:2512.02934 [quant-ph]

    work page arXiv

  22. [22]

    S. Choi, Y. Bao, X.-L. Qi, and E. Altman, Quantum error correction in scrambling dynamics and measurement- induced phase transition, Phys. Rev. Lett.125, 030505 (2020)

    2020

  23. [23]

    M. J. Gullans and D. A. Huse, Dynamical purification phase transition induced by quantum measurements, Phys. Rev. X10, 041020 (2020)

    2020

  24. [24]

    González-García, S

    S. González-García, S. Sang, T. H. Hsieh, S. Boixo, G. Vidal, A. C. Potter, and R. Vasseur, Random insights into the complexity of two-dimensional tensor network calculations, Phys. Rev. B109, 235102 (2024)

    2024

  25. [25]

    J. Chen, J. Jiang, D. Hangleiter, and N. Schuch, Sign problem in tensor-network contraction, PRX Quantum 6, 010312 (2025)

    2025

  26. [26]

    Jiang, J

    J. Jiang, J. Chen, N. Schuch, and D. Hangleiter, Positive bias makes tensor-network contraction tractable, inProceedings of the 57th Annual ACM Symposium on Theory of Computing, STOC ’25 (Association for Computing Machinery, New York, NY, USA, 2025) p. 471–482

    2025

  27. [27]

    Raussendorf and H

    R. Raussendorf and H. J. Briegel, A one-way quantum computer, Phys. Rev. Lett.86, 5188 (2001)

    2001

  28. [28]

    Gross, J

    D. Gross, J. Eisert, N. Schuch, and D. Perez-Garcia, Measurement-based quantum computation beyond the one-way model, Phys. Rev. A76, 052315 (2007)

    2007

  29. [29]

    Aaronson, Quantum computing, postselection, and probabilistic polynomial-time, Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences461, 3473 (2005)

    S. Aaronson, Quantum computing, postselection, and probabilistic polynomial-time, Proceedings of the Royal Society A: Mathematical, Physical and Engineering Sciences461, 3473 (2005)

    2005

  30. [30]

    See Supplemental Material, which contains Refs. [21, 32, 48], for derivation of the cluster-state amplitude and mapping to transfer matrix, additional entanglement and spectral results of translation-invariant model under different boundary conditions, and entanglement and spectral transitions of spatially-disordered model

  31. [31]

    Akila, D

    M. Akila, D. Waltner, B. Gutkin, and T. Guhr, Particle- time duality in the kicked ising spin chain, J. Phys. A 49, 375101 (2016)

    2016

  32. [32]

    Bertini, P

    B. Bertini, P. Kos, and T. Prosen, Exact spectral form factor in a minimal model of many-body quantum chaos, Phys. Rev. Lett.121, 264101 (2018)

    2018

  33. [33]

    Whileforh R = 0, π/2theunitarymodelcanbedescribed in terms of free fermions, forh R =π/4the transfer matrixTis a Clifford unitary

  34. [34]

    Bertini, P

    B. Bertini, P. Kos, and T. Prosen, Entanglement spreading in a minimal model of maximal many-body quantum chaos, Phys. Rev. X9, 021033 (2019)

    2019

  35. [35]

    Piroli, B

    L. Piroli, B. Bertini, J. I. Cirac, and T. Prosen, Exact dynamics in dual-unitary quantum circuits, Phys. Rev. B101, 094304 (2020)

    2020

  36. [36]

    Zhou and A

    T. Zhou and A. W. Harrow, Maximal entanglement velocity implies dual unitarity, Phys. Rev. B106, L201104 (2022)

    2022

  37. [37]

    The various diagnostics provide consistent evidences of the volume-law and area-law behavior athI <∼ 0.1and hI >∼ 0.2, respectively, while none of them gives a sharp estimate of critical pointhI,c of the entanglement transition. Although antipodal mutual information peaks within the intermediate regionh I ∈[0.1,0.2], the finite-size effect and the associa...

  38. [38]

    Vidal, Classical simulation of infinite-size quantum lattice systems in one spatial dimension, Phys

    G. Vidal, Classical simulation of infinite-size quantum lattice systems in one spatial dimension, Phys. Rev. Lett. 98, 070201 (2007)

    2007

  39. [39]

    Y. Li, X. Chen, and M. P. A. Fisher, Measurement-driven entanglement transition in hybrid quantum circuits, Phys. Rev. B100, 134306 (2019)

    2019

  40. [40]

    2(b) is exponentially small inLfor Haar-random states

    The mutual informationIAB in the geometry of Fig. 2(b) is exponentially small inLfor Haar-random states. Such states provide a baseline expectation for behavior in a volume-law phase. 7

  41. [41]

    Lu and T

    T.-C. Lu and T. Grover, Spacetime duality between localization transitions and measurement-induced transitions, PRX Quantum2, 040319 (2021)

    2021

  42. [42]

    L. Su, A. Clerk, and I. Martin, Dynamics and phases of nonunitary floquet transverse-field ising model, Phys. Rev. Res.6, 013131 (2024)

    2024

  43. [43]

    Byun and P

    S.-S. Byun and P. J. Forrester,Progress on the Study of the Ginibre Ensembles, KIAS Springer Series in Mathematics, Vol. 3 (Springer Singapore, 2025)

    2025

  44. [44]

    The scaling ofρ0 −ρe withLis clearer than that ofρ0 −ρ1 because the magnitudes of leading eigenvalues often cross when tensor-network parameters are varied

  45. [45]

    Ippoliti and V

    M. Ippoliti and V. Khemani, Postselection-free entanglement dynamics via spacetime duality, Phys. Rev. Lett.126, 060501 (2021)

    2021

  46. [46]

    Ippoliti, T

    M. Ippoliti, T. Rakovszky, and V. Khemani, Fractal, logarithmic, and volume-law entangled nonthermal steady states via spacetime duality, Phys. Rev. X12, 011045 (2022)

    2022

  47. [47]

    Gopalakrishnan and M

    S. Gopalakrishnan and M. J. Gullans, Entanglement and purification transitions in non-Hermitian quantum mechanics, Phys. Rev. Lett.126, 170503 (2021)

    2021

  48. [48]

    Kawabata, K

    K. Kawabata, K. Shiozaki, M. Ueda, and M. Sato, Symmetry and topology in non-hermitian physics, Phys. Rev. X9, 041015 (2019). END MA TTER Here we investigate the entanglement transition using infinite MPS, and also by computing Rényi entanglement entropies with indexn <1, which characterize the ability of MPS to accurately describe the row state. Followin...

    2019