Pith. sign in

REVIEW 5 major objections 6 minor 34 references

On two-dimensional tensor network group symmetries

T0 review · 5 major / 6 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 2D tensor network symmetry labeled by a nontrivial 4-cocycle cannot have a unique symmetric ground state.

desk verdict Explicit and potentially useful tensor network constructions for 4-cocycle symmetries, but the no-go theorem depends on an underived equation. read the letter →

arxiv 2507.16475 v1 pith:75SE2Z6D submitted 2025-07-22 quant-ph cond-mat.str-el

José Garre-Rubio , András Molnár This is my paper · ORCID
classification quant-phcond-mat.str-el MSC 20J0681P6881T45
keywords tensornetworksprojectedentangledpairstatesgroupcohomology4-cocycleanomalysymmetry-protectedtopologicalphasesmatrixproductoperatorsfinitesymmetryground-statedegeneracy
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

Two-dimensional tensor networks on a hexagonal lattice can represent a finite group $G$ as operators $O_g$, provided the local fusion, orthogonality, and associativity data satisfy consistency equations; the failure of exact associativity is a phase $\omega(g,h,k,l)$ obeying the 4-cocycle equation. The paper's central claim is that when these operators act as symmetries of projected entangled pair states (PEPS), the same 4-cocycle controls the ground-state structure: a unique symmetric ground state forces $\omega$ to be a 4-coboundary, so any nontrivial $\omega$ forces ground-state degeneracy. The same local data are repackaged into unitary tensor networks that act as on-site symmetry operators for $(3+1)$D symmetry-protected topological states, pushing the anomaly to the virtual boundary. This matters because it turns a cohomological classification of anomalous symmetries into explicit local tensor equations that can be checked and used to build states.

What carries the argument

The central object is a fusion matrix product operator built from a five-index tensor $F_{g,h}$, together with associator tensors $R_{g,h,k}$ on a hexagonal-lattice tensor network; local orthogonality $F_{g,h}F_{k,l}=\delta_{g,k}\delta_{h,l}$ and the pentagon equation hold only up to a phase $\omega$, the 4-cocycle. For the PEPS action, the analogous data are action MPOs $W_{g,x}$ and mixed associators $G_{g,h,x}$, whose pentagon identity yields the mixed equation (21). The triple-line and unitary realizations are explicit ways to assign the tensors from a chosen 4-cocycle.

What would settle it

Take $G=\mathbb{Z}_2\times\mathbb{Z}_2$ with the nontrivial 4-cocycle $p_0=p_1=1$ of the paper's example and attempt to build a symmetric PEPS with a unique ground state using local action MPOs obeying the orthogonality condition; Eq. (21) forces that cocycle to be a 4-coboundary, so producing such a state would refute the central claim. A direct numerical construction or search for such a state would therefore settle it.

Watch

Extended reading notes

Core claim

On its own terms, the paper's discovery is the mixed cocycle equation (21), $$\Lambda^x_{g,h,k}\Lambda^x_{g,hk,l}\Lambda^x_{h,k,l} = \Lambda^x_{g,h,kl}\Lambda^x_{gh,k,l}\,\omega(g,h,k,l),$$ which ties the 4-cocycle $\omega$ of the symmetry operator to the associators $\Lambda^x$ of its local action on a PEPS ground state $|\psi_x\rangle$. Setting $|X|=1$ makes $\omega$ a ratio of $\Lambda$'s, i.e. a 4-coboundary, proving that a nontrivial cohomology class forbids a unique symmetric ground state. The paper also constructs explicit realizations, including a unitary one, and shows that the unitary operators give on-site symmetries of a $(3+1)$D GHZ-type state whose virtual boundary carries the 4-cocycle anomaly.

Load-bearing premise

The argument assumes the symmetry acts on the PEPS ground space through a local action MPO $W_{g,x}$ satisfying an orthogonality relation; if a symmetric state admits no such local action, the proof that a nontrivial 4-cocycle forces degeneracy does not go through.

Editorial extensions

If this is right

  • A gapped $(2+1)$D Hamiltonian whose ground state is a PEPS with a local action MPO symmetry and a nontrivial 4-cocycle must have a degenerate ground space, with degeneracy at least $|G/H|$ for the unbroken subgroup $H$.
  • For the unbroken subgroup $H$, Eq. (21) trivializes $\omega$, and the $\Lambda$'s become a 3-cocycle of $H$, so the symmetric phase carries an SPT label in $H^3(H,U(1))$.
  • The equivalence classes of $\Lambda$ under the phase freedom (20) are phase invariants, and two PEPS with the same cocycle and the same $\Lambda$ class are connected by a continuous symmetric path.
  • The unitary tensor network operators give explicit on-site symmetry operators for $(3+1)$D SPT models, with the 4-cocycle anomaly realized at the virtual boundary.
  • For $\omega=1$ the construction reduces to the known tensor network representation of finite group symmetries; the nontrivial $\omega$ adds the anomaly index that controls the ground-state degeneracy.

Reading between the lines

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

  • Beyond the paper's own claims, Eq. (21) can be used as a numerical diagnostic: computing the $\Lambda$ class of any symmetric PEPS with local action MPOs determines whether its symmetry is anomalous without constructing the Hamiltonian.
  • The no-go statement is conditional on the local action MPO orthogonality assumption; the paper does not rule out unique symmetric ground states for anomalous symmetries acting through nonlocal or non-orthogonal actions.
  • The construction suggests a dimensional ladder—MPO symmetries with 3-cocycles organize 2D SPT phases, TNO symmetries with 4-cocycles organize 3D SPT phases—so extending the local equations to fusion 2-category symmetries is a natural next step, one the paper explicitly flags.
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

5 major / 6 minor

Summary. The paper introduces two-dimensional tensor network operators labeled by a finite group G and a 4-cocycle ω ∈ H⁴(G,U(1)), providing two explicit constructions (a non-unitary 'triple-line' representation and a unitary one), and then uses them to study (2+1)D gapped phases with such anomalous symmetries acting on PEPS. The central claim is that a nontrivial 4-cocycle ω forbids a unique symmetric ground state: from the mixed cocycle condition (Eq. 21), the authors derive that |X|=1 forces ω to be a 4-coboundary. They also propose a (3+1)D SPT construction whose boundary anomaly is carried by the 2D tensor network operator, and present a concrete example for G=Z₂×Z₂.

Significance. If the central claims hold, the paper provides a unified tensor-network framework for anomalous 2D group symmetries labeled by H⁴(G,U(1)), with an explicit no-go theorem for unique symmetric ground states and a concrete unitary realization. The explicit local tensor equations, the Z₂×Z₂ example, and the proposed connection to module 2-categories over 2Vec^ω_G are valuable and could serve as a basis for further work. The paper also makes contact with concurrent work on anomaly diagnosis via symmetry restriction, which is useful for positioning. However, the main theorem is not fully supported as written: the key mixed cocycle equation (21) is asserted without derivation, and several other load-bearing statements are only sketched.

major comments (5)
  1. [Sec. 3, Eq. (21)] The mixed 4-cocycle condition Λ^x_{g,h,k}Λ^x_{g,hk,l}Λ^x_{h,k,l} = Λ^x_{g,h,kl}Λ^x_{gh,k,l}ω(g,h,k,l) is stated as 'The associahedron relation implies...' with no derivation. This equation is the sole bridge from the local tensor data to the no-go result that |X|=1 forces ω to be a 4-coboundary. Without an explicit derivation, analogous to the pentagon/associahedron calculation in Sec. 2, the main theorem of the paper is unsupported. Please provide the diagrammatic or algebraic steps leading from the action MPO data to Eq. (21).
  2. [Sec. 2, Eqs. (10) and (12)] The constant ω(g,h,k,l) is initially defined by a diagram that evaluates to ω·D, where D is the product of the dimensions of the virtual spaces (k,l), (h,kl), and (g,hkl). The associators R are only assumed to possess left inverses (Eq. (6)), not to be normalized or unitary. In the standard 4-cocycle consistency condition, ω is a U(1) phase and the dimension factor D must cancel in the pentagon/associahedron relations. The authors do not show this cancellation or impose a normalization condition on R. If D does not cancel, Eq. (12) would be modified, and the claimed H⁴(G,U(1)) classification would not follow. This requires clarification.
  3. [Sec. 3, classification claim] The statement that 'With the same techniques that we developed in Ref. [15], we could show that a differentiable change of the PEPS tensor ... will modify the action MPOs in such a way that the new action associators will define a Λ that belongs to the same class' is an assertion, not a proof. Since this invariance is essential for the claim that equivalence classes of Λ are invariants of the quantum phase, the proof should be included or the claim explicitly restated as a conjecture. The current text leaves the classification result conditional on an unproven lemma.
  4. [Sec. 3, main assumption on action MPO] The no-go theorem and the classification depend on the assumption that the symmetry action on a PEPS ground state is realized by a local action MPO W_{g,x} satisfying the orthogonality condition shown after Eq. (18). This is a substantial assumption: it is not proven that every symmetric PEPS admits such an action MPO, nor is the condition motivated by canonical forms. The paper should either justify this assumption (e.g., by showing it holds for PEPS in a canonical form) or clearly state the scope limitation, since it directly affects the validity of the no-go theorem for arbitrary symmetric PEPS.
  5. [Sec. 4.2, (3+1)D SPT construction] The claim that the on-site unitary u_g in Eq. (26) defines a global symmetry of |Ψ_3D⟩ and that the boundary anomaly is described by the 2D TNO is only sketched: the text says the virtual TNO 'grows correctly when concatenated' and shows several diagrams, but no explicit equations or proofs are provided for the key pulling-through and concatentation steps. As this construction is a stated contribution, the derivation should be made more explicit, or the section should be labeled as a sketch with a clear list of unproven steps.
minor comments (6)
  1. [Sec. 3, text after Eq. (19)] There is a typo: 'developped' should be 'developed'.
  2. [Sec. 2.2, Eq. (16)] The sentence 'which holds due to the 4-cocycle equation' is followed by an equation whose indexing is hard to follow; writing out the substitutions explicitly would improve readability.
  3. [Sec. 2, Eqs. (10)-(12)] The same symbol ω is used for the constant in Eq. (10) and for the 4-cocycle in Eq. (12); if these differ by dimension factors, the notation is confusing. Please clarify by using a different symbol for the dimension-dependent constant or by fixing a normalization.
  4. [Sec. 4.1] The review of (2+1)D SPT phases is longer than necessary relative to the new material; condensing it would help the reader focus on the new 3D construction.
  5. [Sec. 2.2, Fig. 1] The description of the unitary tensor network relies heavily on Fig. 1 and several other diagrams that are not fully specified in the text. A more detailed explanation of the tensor indices and bond dimensions would make the construction easier to verify.
  6. [References] Ref. [27] lists the title with a typo ('Projeted' instead of 'Projected'). Also, the paper should discuss in more detail the relationship with the concurrent work Ref. [23], beyond the 'Noted Added' sentence.

Circularity Check

0 steps flagged · score 0.0 of 10

No circularity: the derivation chain is a genuine mathematical construction; Eq. (21) is a consistency condition, not a restatement of the input, and no fitted parameter is renamed as a prediction.

full rationale

I walked the derivation chain. Section 2 constructs a tensor-network representation of G from local fusion tensors F and associators R; the constant ω is introduced in Eq. (10) as the proportionality factor forced by the pentagon-type move, and Eq. (12) is stated as the resulting 4-cocycle condition, with the associahedron coherence checked against Kitaev's Fig. 18. This is a derivation, not an input. The triple-line and unitary realizations then assign concrete tensor values using the 4-cocycle and verify that the local equations hold; this is construction-by-design, but it is not circular because the target group representation is not assumed—it is assembled and checked. Section 3 assumes the existence of a local action MPO W_{g,x} and mixed associators G; the mixed constant Λ is defined by Eq. (19) from the pentagon relation, and Eq. (21) is the resulting mixed 4-cocycle consistency condition. The no-go statement that |X|=1 forces ω to be a 4-coboundary is a direct algebraic consequence of Eq. (21). No parameter was fitted to the conclusion, and no 'prediction' is simply the input re-labeled. The skeptic's point that Eq. (21) is asserted without a displayed derivation is a rigor or verification gap, not a circular reduction; the equation has independent mathematical content (it constrains Λ relative to ω and matches the module 2-category interpretation). The self-citation to Ref. [15] is used only as 'same techniques' for a phase-invariance argument, not as the load-bearing premise for the central anomaly result. Therefore no significant circularity is present.

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

The paper introduces no fitted parameters. The constructions are parameterized by an input 4-cocycle ω. The main assumptions are the existence of associators and the local MPO form of symmetry actions on PEPS. These are domain assumptions common in tensor network analyses of symmetries.

assumptions (4)
  • ad hoc to paper There exist associator tensors R_{g,h,k} with left inverses satisfying Eq. (7).
    Introduced in Sec. 2 as 'our most important assumption'. It is needed to derive the pentagon relation and the 4-cocycle equation for the tensor network representation.
  • domain assumption The tensors T_g are left-invertible.
    Assumed in Sec. 2 to cancel tensors in diagrammatic proofs, e.g., in deriving associativity.
  • domain assumption The symmetry action on PEPS ground states is realized by a local action MPO W_{g,x} satisfying an orthogonality condition.
    Assumed in Sec. 3 to derive the mixed cocycle equation (21), which is the basis of the phase classification.
  • domain assumption The PEPS tensors A_x are left-invertible.
    Used in Sec. 3 to derive the pentagon relation involving action associators.

how reviews work

0 comments
Cite this review

Pith. "Pith review of On two-dimensional tensor network group symmetries." pith.science (2026). https://pith.science/paper/75SE2Z6D

@misc{pith2026250716475,
  author       = {Pith},
  title        = {Pith review of: On two-dimensional tensor network group symmetries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/75SE2Z6D}},
  note         = {Machine review of arXiv:2507.16475}
}
read the original abstract

We introduce two-dimensional tensor network representations of finite groups carrying a 4-cocycle index. We characterize the associated gapped (2+1)D phases that emerge when these anomalous symmetries act on tensor network ground states. We further develop related tensor network unitaries that generate symmetric states representing (3+1)D symmetry protected topological phases. Although aspects of these constructions have been previously addressed, our contribution unifies them within a single tensor network framework and emphasizes the explicit formulation of local tensor equations encoding global consistency conditions.

Figures

Figures reproduced from arXiv: 2507.16475 by the authors.

Figure 1
Figure 1. The tensor network unitary representation of a finite gr [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

34 extracted references · 31 canonical work pages

  1. [15]

    Quantum, 7:927, February 2023

    Jos´ e Garre-Rubio, Laurens Lootens, and Andr´ as Moln´ ar.Classifying phases protected by matrix product operator symmetries using matrix product states. Quantum, 7:927, February 2023

    work page 2023

  2. [1]

    Ignacio Cirac, David P´ erez-Garc ´ ıa, Norbert Schuch, and F rank Verstraete

    J. Ignacio Cirac, David P´ erez-Garc ´ ıa, Norbert Schuch, and F rank Verstraete. Matrix product states and projected entangled pair states: Concepts, symmet ries, theorems. Reviews of Modern Physics, 93(4), dec 2021

    work page 2021

  3. [2]

    Matrix product operator symmetries and intertwiners in string-ne ts with domain walls

    Laurens Lootens, Jurgen Fuchs, Jutho Haegeman, Christoph Schweigert, and Frank Verstraete. Matrix product operator symmetries and intertwiners in string-ne ts with domain walls. SciPost Phys., 10:53, 2021

    work page 2021

  4. [3]

    Ignacio Cirac, and David P´ erez-Garc ´ ıa

    Andras Molnar, Alberto Ruiz de Alarc´ on, Jos´ e Garre-Rubio, Norbert Schuch, J. Ignacio Cirac, and David P´ erez-Garc ´ ıa. Matrix product operator algebras i: representations of weak hopf algebras and projected entangled pair states, 2022

    work page 2022

  5. [4]

    Tensor network approach to electromagnetic duality in (3+1)d topological gauge models

    Clement Delcamp. Tensor network approach to electromagnetic duality in (3+1)d topological gauge models. Journal of High Energy Physics , 2022(8), August 2022

    work page 2022

  6. [5]

    M. B. Hastings and Xiao-Gang Wen. Quasiadiabatic continuation of quantum states: The stability of topological ground-state degeneracy and emergent gauge inv ariance. Phys. Rev. B , 72:045141, Jul 2005

    work page 2005

  7. [6]

    Fannes, B

    M. Fannes, B. Nachtergaele, and R. F. Werner. Finitely correlat ed states on quantum spin chains. Communications in Mathematical Physics , 144(3):443–490, Mar 1992

    work page 1992

  8. [7]

    M.Wolf D

    M. M.Wolf D. P´ erez-Garc ´ ıa, F. Verstraete and J. I. Cirac. Matrix product state representations. Quant. Inf. Comput. , 7(401), 2007

    work page 2007

Show all 34 references
  1. [8]

    Turner, Erez Berg, and Masaki Oshikawa

    Frank Pollmann, Ari M. Turner, Erez Berg, and Masaki Oshikawa . Entanglement spectrum of a topological phase in one dimension. Phys. Rev. B , 81:064439, Feb 2010

    2010

  2. [9]

    Classification of g apped symmetric phases in one-dimensional spin systems

    Xie Chen, Zheng-Cheng Gu, and Xiao-Gang Wen. Classification of g apped symmetric phases in one-dimensional spin systems. Phys. Rev. B , 83:035107, Jan 2011

    2011

  3. [10]

    Classifying quantum phases using matrix product states and projected entangled pair states

    Norbert Schuch, David P´ erez-Garc ´ ıa, and Ignacio Cirac. Classifying quantum phases using matrix product states and projected entangled pair states. Phys. Rev. B , 84:165139, Oct 2011

    2011

  4. [11]

    Classification of symmetry protected topologic al phases in quantum spin chains, 2021

    Yoshiko Ogata. Classification of symmetry protected topologic al phases in quantum spin chains, 2021

    2021

  5. [12]

    S ymmetry protected topological orders and the group cohomology of their symmetry group

    Xie Chen, Zheng-Cheng Gu, Zheng-Xin Liu, and Xiao-Gang Wen. S ymmetry protected topological orders and the group cohomology of their symmetry group. Phys. Rev. B , 87:155114, Apr 2013

    2013

  6. [13]

    Anomalies of discrete sym metries in various dimensions and group cohomology, 2014

    Anton Kapustin and Ryan Thorngren. Anomalies of discrete sym metries in various dimensions and group cohomology, 2014

    2014

  7. [14]

    Two-dimensional sy mmetry-protected topological orders and their protected gapless edge excitations

    Xie Chen, Zheng-Xin Liu, and Xiao-Gang Wen. Two-dimensional sy mmetry-protected topological orders and their protected gapless edge excitations. Phys. Rev. B , 84:235141, Dec 2011

    2011

  8. [16]

    Generalized global symme- tries

    Davide Gaiotto, Anton Kapustin, Nathan Seiberg, and Brian Willet t. Generalized global symme- tries. Journal of High Energy Physics , 2015(2), February 2015

    2015

  9. [17]

    Fusion category symmetry i: Anom aly in-flow and gapped phases

    Yifan Wang Ryan Thorngren. Fusion category symmetry i: Anom aly in-flow and gapped phases. arXiv:1912.02817, 2019

    1912 arXiv

  10. [18]

    (2+1)d lattice models and tensor networks for gapped phases with categorical s ymmetry, 2025

    Kansei Inamura, Sheng-Jie Huang, Apoorv Tiwari, and Sakura Schafer-Nameki. (2+1)d lattice models and tensor networks for gapped phases with categorical s ymmetry, 2025. 17

    2025

  11. [19]

    Gapped phases in (2+1)d with non-invertible symmetrie s: Part i, 2024

    Lakshya Bhardwaj, Daniel Pajer, Sakura Schafer-Nameki, A poorv Tiwari, Alison Warman, and Jingxiang Wu. Gapped phases in (2+1)d with non-invertible symmetrie s: Part i, 2024

    2024

  12. [20]

    Gapped phases in (2+1)d with non-invertible symmetries: Part ii, 2025

    Lakshya Bhardwaj, Sakura Schafer-Nameki, Apoorv Tiwari, a nd Alison Warman. Gapped phases in (2+1)d with non-invertible symmetries: Part ii, 2025

    2025

  13. [21]

    Higher categorical symmetr ies and gauging in two- dimensional spin systems

    Clement Delcamp and Apoorv Tiwari. Higher categorical symmetr ies and gauging in two- dimensional spin systems. SciPost Physics , 16(4), April 2024

    2024

  14. [22]

    Fusion surface models: 2 +1d lattice models from fusion 2-categories

    Kansei Inamura and Kantaro Ohmori. Fusion surface models: 2 +1d lattice models from fusion 2-categories. SciPost Physics , 16(6), June 2024

    2024

  15. [23]

    Anomaly diagnosis via symmetry re striction in two-dimensional lattice systems

    Kyle Kawagoe and Wilbur Shirley. Anomaly diagnosis via symmetry re striction in two-dimensional lattice systems. preprint arxiv:2507.07430, 2025

    2025 arXiv

  16. [24]

    Anyons in an exactly solved model and beyond

    Alexei Kitaev. Anyons in an exactly solved model and beyond. Annals of Physics , 321(1):2 – 111,

  17. [25]

    Bultinck, M

    N. Bultinck, M. Mari¨ en, D.J. Williamson, M.B. S ¸ahino˘ glu, J. Haeg eman, and F. Verstraete. Anyons and matrix product operator algebras. Annals of Physics , 378:183–233, 2017

    2017

  18. [26]

    Wang and Xiao-Gang Wen

    Juven C. Wang and Xiao-Gang Wen. Non-abelian string and partic le braiding in topological order: Modular SL(3 , Z) representation and (3 + 1)-dimensional twisted gauge theory. Phys. Rev. B, 91:035134, Jan 2015

    2015

  19. [27]

    Verstraete and J

    F. Verstraete and J. I. Cirac. Renormalization algorithms for q uantum-many body systems in two and higher dimensions. ArXiv: cond-mat/0407066

    arXiv

  20. [28]

    Peps as ground states: Degeneracy and topology

    Norbert Schuch, Ignacio Cirac, and David Perez-Garcia. Peps as ground states: Degeneracy and topology. Annals of Physics , 325(10):2153 – 2192, 2010

    2010

  21. [29]

    On tensor network repr esentations of the (3+1)d toric code

    Clement Delcamp and Norbert Schuch. On tensor network repr esentations of the (3+1)d toric code. Quantum, 5:604, December 2021

    2021

  22. [30]

    Williamson, Clement Delcamp, Frank Verstraete, and No rbert Schuch

    Dominic J. Williamson, Clement Delcamp, Frank Verstraete, and No rbert Schuch. On the stability of topological order in tensor network states. Physical Review B , 104(23), December 2021

    2021

  23. [31]

    Higher berry phase from proj ected entangled pair states in (2 + 1) dimensions

    Shuhei Ohyama and Shinsei Ryu. Higher berry phase from proj ected entangled pair states in (2 + 1) dimensions. Phys. Rev. B , 111:045112, Jan 2025

    2025

  24. [32]

    1+1d spt phases with fus ion category symmetry: interface modes and non-abelian thouless pump, 2024

    Kansei Inamura and Shuhei Ohyama. 1+1d spt phases with fus ion category symmetry: interface modes and non-abelian thouless pump, 2024

    2024

  25. [33]

    Fractional domain wa ll statistics in spin chains with anomalous symmetries

    Jos´ e Garre-Rubio and Norbert Schuch. Fractional domain wa ll statistics in spin chains with anomalous symmetries. SciPost Physics , 18(2), February 2025. 18

    2025

  26. [2006]

    January Special Issue

Pith tools

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