Pith. sign in

REVIEW 3 major objections 4 minor 45 references

Generalised ansatz for continuous Matrix Product States

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

Pith's one-line read A generalized cMPS ansatz expresses the continuum limit of every MPS.

desk verdict A clean and useful transfer-matrix construction that plausibly closes a known representability gap, but the paper should say more precisely what it means for a generalized cMPS to 'represent' a continuum limit state. read the letter →

arxiv 1908.09761 v2 pith:OG2SATBH submitted 2019-08-26 quant-ph

classification quant-ph MSC 81P45
keywords continuousmatrixproductstatescontinuumlimitinfinitelydivisiblequantumchannelsprojectorLindbladLiouvilliantensornetworkssuperselectionsectors
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

Continuous matrix product states (cMPS) miss the continuum limits of many matrix product states (MPS), because a cMPS transfer matrix is always a Markovian channel $e^L$, whereas an MPS with a continuum limit has transfer matrix $P e^{aL}$ with $P$ a projector quantum channel. This paper claims that the missing projector can be supplied by a generalized ansatz: a finite sum of cMPS with boundary operators given by the Kraus operators of $P$, each attached to an ancilla state. The ansatz reproduces the transfer matrix $P e^{aL}$ of any MPS continuum limit, and is interpreted as a standard cMPS concatenated with an element at the closure of the set of cMPS that supplies $P$. If correct, it gives every translationally invariant MPS with a continuum limit an exact matrix-product description directly in the continuum.

What carries the argument

The load-bearing object is the generalized cMPS ansatz of Eq. (24): a sum over the Kraus operators $\{B_i\}$ of the projector quantum channel $P$, with each $B_i$ used as the boundary operator of a cMPS sharing the same $Q$ and $\{R_\alpha\}$, and each term carrying a distinct ancilla state $|v_i\rangle\in C^K$. The argument turns on the transfer-matrix identity $(\Phi_R,\Phi_R)=P e^{aL}$, because the defining property of an MPS continuum limit is precisely the factorization $E_a=P e^{aL}$ with $PL=PLP$. The condition $PL=PLP$ ensures consistency $\bigl(P e^{aL}\bigr)^N=P e^{NaL}$, so every point of the segment is well defined. Proposition 1 supplies the physical picture: $P$ is a limit point of cMPS transfer matrices $e^{t\tilde L}$ built from explicit jump operators, making the generalized ansatz a concatenation of a closure-of-cMPS object and a standard cMPS.

What would settle it

Take the MPS family $|0\cdots0\rangle+|1\cdots1\rangle$, whose transfer matrix is $P$ alone, and compute its continuum limit directly by applying the p-refinement isometries. Compare that state with the generalized cMPS states obtained from two different Kraus decompositions of $P$; Example 1 gives different states, such as $(|v_0\rangle+|v_1\rangle)\otimes|\Omega\rangle$ and $|w_0\rangle\otimes|\Omega\rangle$. If the refined limit disagrees with one of these representations in a global or ancilla-sensitive observable, then transfer-matrix equality does not uniquely determine the continuum limit state and the state-level reading of Theorem 4 would be refuted.

Watch

Extended reading notes

Core claim

Theorem 4 is the central claim. Let $\mathcal V(A)=\{|V_N(A)\rangle\}$ be a translationally invariant MPS family with a continuum limit in the sense of Definition 2, so its transfer matrix is $E_a=P e^{aL[Q,\{R_\alpha\}]}$ with $P^2=P$ a projector quantum channel, $L[Q,\{R_\alpha\}]$ a Lindblad Liouvillian, and $PL=PLP$. If $P=\sum_{i=1}^K B_i\otimes \bar B_i$ is a Kraus decomposition of $P$, then the continuum limit state is represented by the generalized cMPS $|\Phi_R[\{v_i\},\{B_i\},Q,\{R_\alpha\}]\rangle=\sum_{i=1}^K |v_i\rangle\otimes |\varphi_R[B_i,Q,\{R_\alpha\}]\rangle$, where $\{|v_i\rangle\}$ is an orthonormal basis of the ancilla space $C^K$. The proof evaluates the transfer matrix $(\Phi_R,\Phi_R)=\bigl(\sum_i B_i\otimes\bar B_i\bigr)e^{aL}=P e^{aL}$, matching the continuum limit. A further result, Proposition 1, shows that every projector channel $P$ equals $\lim_{t\to\infty} e^{t\tilde L}$ for an explicit Lindblad generator, so the projector part of the ansatz can be seen as a cMPS in the thermodynamic limit or with unbounded matrix norm.

Load-bearing premise

The proof shows the generalized cMPS has the same transfer matrix as the continuum limit, then assumes that matching transfer matrices is enough to identify the state; that identification is carried over from earlier work and is not proven in this paper.

Editorial extensions

If this is right

  • Every translationally invariant MPS that has a continuum limit receives an exact continuum matrix-product description, closing the expressivity gap that plain cMPS left open.
  • The correlation-function calculus of cMPS carries over with one replacement: the boundary term $B\otimes\bar B$ becomes the projector $P$, so existing cMPS computational tools apply almost unchanged.
  • The projector's zeros, which act as superselection rules, can be imposed directly at the continuum level instead of by sending matrix norms or system sizes to infinity.
  • The ansatz gives a variational family for continuum theories with several superselection sectors, illustrated by superpositions of ferromagnetic states, the completely depolarizing channel, and the bracket state.

Reading between the lines

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

  • The proof establishes equality of transfer matrices; the step from transfer-matrix equality to equality of states is imported from the continuum-limit framework of earlier work and is not re-proved here, so Theorem 4 is strictly a statement about transfer matrices and quantities determined by them unless that premise holds.
  • A narrower ansatz with the condition $PL=PLP$ built into the state structure may exist; the authors connect this to G-injective MPS, where superpositions with different boundary conditions reproduce the ground-state degeneracy of a parent Hamiltonian, hinting at a continuum parent Hamiltonian problem.
  • The physical role of the ancilla space $C^K$ is left open; identifying an observable that reads the ancilla would turn the generalized ansatz from a mathematical representation into a usable physical variational class.
  • One concrete stress test is to compare the continuum limit obtained by explicit p-refinement with the generalized cMPS for a case where $L=0$, such as the ferromagnetic superposition; any difference in global coherence between the two sectors would show that transfer-matrix matching alone does not fix the state.
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

3 major / 4 minor

Summary. The paper proposes a generalized continuous Matrix Product State (cMPS) ansatz that is claimed to be able to express the continuum limit of any translationally invariant MPS with a continuum limit. The ansatz is a sum of cMPS, each with a different boundary operator given by a Kraus operator of the projector quantum channel P appearing in the transfer matrix E_a = P e^{aL}, together with an ancilla label. The main result, Theorem 4, is that the transfer matrix of this generalized cMPS is exactly P e^{|R|L}, matching the transfer matrix of the MPS continuum limit. The paper also interprets the ansatz as a concatenation of a state at the closure of the set of cMPS and a standard cMPS, and illustrates the construction with several examples, including superpositions of ferromagnetic states, the completely depolarizing channel, and a 'bracket state'.

Significance. If the representability claim is established at the level of states, this would be a clean resolution of a limitation of cMPS identified in Ref. [19]: the missing projector P in the transfer matrix. The construction is elegant and the transfer-matrix computation in Eq. (27) is direct and correct. The paper also gives an explicit construction (Proposition 1) showing that every projector quantum channel arises as the infinite-time limit of a Markovian channel, which is a useful technical result. However, the central gap between transfer-matrix equality and state-level representability currently prevents Theorem 4 from being fully convincing.

major comments (3)
  1. [Section 5.2, Theorem 4 and Eq. (27)] The proof of Theorem 4 establishes only that the transfer matrix of the constructed state |Φ_R> equals P e^{|R|L}. The theorem, however, claims that the continuum limit state of |V_N(A)> is represented by |Φ_R>. Section 5.3 explicitly acknowledges that many continuum states share a given transfer matrix ('there be many states in the continuum whose transfer matrix is E_{|R|}') and then characterizes freedom in P, L, Kraus operators, the ancilla basis, and the Fock space, but it does not provide a state-level identification. To make the theorem sound, the authors must either define precisely what 'represented' means (e.g., equality of all quasi-local correlation functions, or convergence under the refining isometries of Definition 2) and prove that the generalized cMPS satisfies this definition, or restrict the theorem's statement to a transfer-matrix-level statement. Without this, the central claim is not proven.
  2. [Section 5.3 and Example 1] The ambiguity between transfer-matrix and state-level equivalence is concretely visible in Example 1. For the superposition of ferromagnetic states, the generalized cMPS is |Φ_R> = (|v0> + |v1>) ⊗ |Ω_R>. If the ancilla is traced out, the physical state is the vacuum, not a coherent global superposition; if the ancilla is kept, the state lives in C^K ⊗ H_R and no embedding of this space into the original physical space H_R is specified. The paper leaves the physical interpretation of the ancilla as an open question, which is acceptable, but the statement of Theorem 4 requires a clear specification of the sense in which the ancilla-carrying state represents the original MPS continuum limit.
  3. [Section 5.2 and 5.3, N-dependence] Theorem 4 states that 'for any N' the continuum limit state of |V_N(A)> can be represented by the generalized cMPS. The proof, however, only verifies the transfer matrix on a segment of length a, and then uses the consistency condition (E_a)^N = E_{Na} to extend to longer segments. This establishes translational consistency of the transfer matrix but does not show that the sequence of generalized cMPS states, as N varies, is related by the p-refinement and blocking operations that define the continuum limit in Definition 2. Please clarify how the N-dependence of |Φ_R> is meant to match the definition of continuum limit.
minor comments (4)
  1. [Eqs. (24) and (25)] The notation switches between |φ_R[B_i,...]> (physical state after tracing the auxiliary space, used in Eq. (24)) and φ_R[B_i,...] (operator with open auxiliary indices, used in Eq. (25)) without an explicit reminder. Please introduce a clear notational distinction, as this is a common source of confusion in cMPS calculations.
  2. [Section 5.4] The sentence 'the first element can be seen as the thermodynamic limit of a cMPS in the thermodynamic limit' is redundant; one of the two occurrences should be removed.
  3. [Eq. (27)] The inner product in Eq. (26) is defined on (C^K ⊗ M_D ⊗ H_R) × (C^K ⊗ M_D ⊗ H_R), but the state |Φ_R> in Eq. (24) is an element of C^K ⊗ H_R after the auxiliary trace. The proof implicitly uses the open-index version Φ_R of Eq. (25). Please state explicitly that the transfer matrix is defined using the open-index object, as is standard for cMPS.
  4. [Appendix A, proof of Proposition 1] The proof of Proposition 1 treats three special cases and then says that 'putting these three building blocks together' gives the general case. This is plausible, but the text would benefit from a short explicit statement of why the cases (i)-(iii) suffice to cover the general form of P in Eq. (9), especially regarding the interplay between the blocks π_k and the isometries V_k.

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: the generalized cMPS ansatz is constructed from the Kraus operators of P, and Eq. (27) verifies by direct computation that its transfer matrix equals P e^{|R|L}; the state-level gap noted in Sec. 5.3 is a proof limitation, not a circular reduction.

full rationale

The central construction (Theorem 4) is not circular. The generalized cMPS is defined by attaching cMPS states with boundary operators B_i, the Kraus operators of P, to orthonormal ancilla states; the proof computes the transfer matrix directly in Eq. (27) and obtains P e^{|R|L}. No parameter is varied or fitted to force this agreement, and the equality is not assumed but derived from the orthonormality of the ancilla basis and the Kraus decomposition. The paper does rely on the prior characterization of infinitely divisible channels and of MPS continuum limits (Theorem 2 and Theorem 3, from Ref. [19], whose first author is one of the present authors). That reliance is a normal use of an established mathematical result, not a self-citation chain that substitutes for an argument, and the target claim (the new ancilla-sum ansatz) is not identical to that characterization. Section 5.3 explicitly concedes that many continuum states share the same transfer matrix, so the proof establishes transfer-matrix and correlation-function agreement rather than a full state-level identification; this is a limitation of the state-representability claim and a potential correctness concern, but it is not circularity, because the ansatz is not defined in terms of the conclusion and the conclusion does not reduce to the input by construction. Score 1 reflects the minor self-citation dependence; no circular step is identified.

Assumptions & free parameters 0 free parameters · 4 assumptions · 1 invented entities

No parameters are fitted to data anywhere in the paper; the variational data (Q, {R_alpha}, {B_i}, {|v_i>}) are constructed from the given transfer matrix E_a = Pe^{aL}, and the gauge freedom in choosing them is characterized in Section 5.3 (unitary freedom among minimal Kraus decompositions; Liouvillian freedom per [23, Prop. 7.4]). The theta_{k,i} of Eq. (33) encode the fixed-point states sigma_k of P rather than being free settings. The genuinely new structural ingredient is the ancilla register C^K labeling the Kraus boundary operators, whose physical meaning the paper leaves open in Section 6. The claims rest on the external results listed as axioms, most notably Ref. [19]'s characterization of infinitely divisible channels, a published theorem co-authored by one of the present authors, and on the standard channel facts of [23].

assumptions (4)
  • domain assumption Definition 2 (continuum limit of an MPS): the limit is defined via infinite p-refinement together with a regularisation condition (n_k / p^{l_k}) -> 0, following Ref. [19]. Theorem 4 is stated relative to this exact definition.
    The abstract and Section 3.2 state that all claims are relative to this specific notion of continuum limit; under a different definition of the limit, the characterization and the ansatz would not necessarily apply.
  • standard math Theorem 2 of this paper (from Ref. [19]): a quantum channel E is infinitely divisible iff E = P e^L with P a projector quantum channel, L a Lindblad Liouvillian, and PL = PLP; combined with Theorem 3 this characterizes MPS families that have a continuum limit.
    This published characterization is the bridge that turns the MPS continuum limit question into the transfer matrix factorization problem; the ansatz of Eq. (24) is engineered to realize exactly Pe^{aL}. It is cited from [19] and used as a black box.
  • standard math Theorem 1 and Propositions 7.4, 7.5 of Wolf's lecture notes [23]: structure of fixed point sets of quantum channels and freedom in Lindblad representations, used for the decomposition of projector channels and Liouvillian equivalences.
    Section 2.3 uses the fixed point decomposition (3) to characterize projector channels as pinching-plus-depolarising blocks, and Sections 5.3 and Appendix A use the equivalence results for Liouvillians.
  • domain assumption Regularity conditions of Ref. [14, Sec. III] on Q and {R_alpha} that guarantee finite kinetic energy; Section 5.3 asserts the generalized cMPS inherits them from the underlying cMPS.
    The claim that the ansatz is a legitimate physical state in the Fock space with finite kinetic energy depends on these regularity conditions holding for the chosen Q and {R_alpha}.
invented entities (1)
  • Ancilla register C^K attached to the generalized cMPS
    purpose: Labels the K summands, each a cMPS with boundary operator B_i (a Kraus operator of the projector P), so the ansatz of Eq. (24) is a sum of cMPS with different boundary conditions.
    The paper itself lists the physical nature of the ancilla Hilbert space as an open question (Section 6). No observable accessing the ancilla is proposed and no falsifiable handle on it is given; its role is structural within the construction, so independent_evidence is false.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Generalised ansatz for continuous Matrix Product States." pith.science (2026). https://pith.science/paper/OG2SATBH

@misc{pith2026190809761,
  author       = {Pith},
  title        = {Pith review of: Generalised ansatz for continuous Matrix Product States},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/OG2SATBH}},
  note         = {Machine review of arXiv:1908.09761}
}
read the original abstract

Recently it was shown that continuous Matrix Product States (cMPS) cannot express the continuum limit state of any Matrix Product State (MPS), according to a certain natural definition of the latter. The missing element is a projector in the transfer matrix of the MPS. Here we provide a generalised ansatz of cMPS that is capable of expressing the continuum limit of any MPS. It consists of a sum of cMPS with different boundary conditions, each attached to an ancilla state. This new ansatz can be interpreted as the concatenation of a state which is at the closure of the set of cMPS together with a standard cMPS. The former can be seen as a cMPS in the thermodynamic limit, or with matrices of unbounded norm. We provide several examples and discuss the result.

Figures

Figures reproduced from arXiv: 1908.09761 by the authors.

Figure 1
Figure 1. FIG. 1: (Top) Graphical representation of the MPS [PITH_FULL_IMAGE:figures/full_fig_p005_1.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

45 extracted references · 32 canonical work pages

  1. [19]

    Verstraete and J

    F. Verstraete and J. I. Cirac, arXiv:cond-mat/0407066 (2004)

  2. [1]

    Generalised ansatz for continuous Matrix Product States

    INTRODUCTION Tensor networks are a powerful ansatz to describe and sim- ulate quantum many-body systems in an efficient way [1, 2], which originated in the context of condensed matter physics, but has nowadays percolated to other areas—for a recent re- view see [2]. The class of Matrix Product States (MPS) has been extremely successful in capturing ground s...

  3. [2]

    PRELIMINARIES ON QUANTUM CHANNELS In this section we present the background on quantum chan- nels. We will first present the basic definitions (Section 2.1), then characterise Markovian channels (Section 2.2), projector quantum channels (Section 2.3), and finally infinitely divisible channels (Section 2.4). 2.1. Basic definitions Throughout this paper we letMD...

  4. [3]

    3.E =PeL whereP is a projector quantum channel andL is a Liouvillian of Lindblad form such thatPL =PLP

    There is an integer p > 1 such thatE = (Ep𝓁)p𝓁 for all 𝓁 ∈ N, and for all sequences {nk,𝓁 k}∞ k=1 fulfilling limk→∞ nk/p𝓁k = 0, (Enk)p𝓁k →P , whereP is a projector quantum channel. 3.E =PeL whereP is a projector quantum channel andL is a Liouvillian of Lindblad form such thatPL =PLP. Therefore, Markovian quantum channels are infinitely divisible channels wi...

  5. [4]

    We will first present the basic definitions of MPS (Sec- tion 3.1) and then review the results on their continuum limits (Section 3.2)

    PRELIMINARIES ON MATRIX PRODUCT STATES In this section we present the background on Matrix Prod- uct States (MPS) and their continuum limits required for this paper. We will first present the basic definitions of MPS (Sec- tion 3.1) and then review the results on their continuum limits (Section 3.2). 3.1. Matrix Product States In this work we consider exclu...

  6. [5]

    open auxiliary indices

    PRELIMINARIES ON CONTINUOUS MATRIX PRODUCT STATES In this section we present the background on cMPS, mainly by following Ref. [14]. We will first present the mathematical setting (Section 4.1) and then the definition of continuous MPS (Section 4.2). 4.1. Mathematical setting We first define a segment R as an interval of the reals R = [x, y], where x < y and x...

  7. [6]

    REPRESENTING THE CONTINUUM LIMIT OF AN MPS In this section we present the core of this work, namely a generalised ansatz of cMPS. First we will precisely state the problem (Section 5.1), then we will present a generalised ansatz of cMPS addressing this problem (Section 5.2), we will discuss it (Section 5.3), provide an interpretation thereof (Section 5.4)...

  8. [7]

    A Liouvillian of Lindblad form L such that PL = PLP and such that Ea = PeaL

Show all 45 references
  1. [8]

    The Kraus operators{Bi}K i=1 of P

  2. [9]

    Q [satisfying (2)] and{R}q α=1 such that L = Q⊗ I + I⊗ ¯Q + ∑q α=1 Rα⊗ ¯Rα

  3. [10]

    An orthonormal basis{|vi⟩}K i=1 of CK

  4. [11]

    A second comment concerns Definition 2, i.e

    A Fock space HR. A second comment concerns Definition 2, i.e. the definition of continuum limit. As mentioned above, the equal superpo- sition of ferromagnetic states|0, 0,..., 0⟩ +|1, 1,..., 1⟩ has a continuum limit. However, this state is distinguishable from a probabilistic m...

  5. [12]

    Dok- toratsstipendium aus der Nachwuchsf¨orderung

    CONCLUSIONS AND OUTLOOK In summary, we have proposed a generalised ansatz of con- tinuous MPS which can express the continuum limit of any MPS (Theorem 4), according to the definition of continuum limit given in Definition 2. The ansatz consists of a sum of cMPS, where each cMPS...

  6. [13]

    Or ´us, Ann

    R. Or ´us, Ann. Phys. 349, 117 (2014)

  7. [14]

    Orus, Nat

    R. Orus, Nat. Rev. Phys. 1, 538 (2019)

  8. [15]

    Fannes, B

    M. Fannes, B. Nachtergaele, and R. F. Werner, Commun. Math. Phys. 144, 443 (1992)

  9. [16]

    Vidal, Phys

    G. Vidal, Phys. Rev. Lett. 91, 147902 (2003)

  10. [17]

    Schollw ¨ock, Annals of Physics 326, 96 (2011)

    U. Schollw ¨ock, Annals of Physics 326, 96 (2011)

  11. [18]

    Perez-Garcia, F

    D. Perez-Garcia, F. Verstraete, M. M. Wolf, and J. I. Cirac, Quantum Inf. Comput. 7, 401 (2007)

  12. [20]

    Verstraete, J

    F. Verstraete, J. J. Garcia-Ripoll, and J. I. Cirac, Phys. Rev. Lett. 93, 207204 (2004)

  13. [21]

    Zwolak and G

    M. Zwolak and G. Vidal, Phys. Rev. Lett. 93, 207205 (2004)

  14. [22]

    C. V . Kraus, N. Schuch, F. Verstraete, and J. I. Cirac, Phys. Rev. A 81, 52338 (2010)

  15. [23]

    J. I. Cirac and G. Sierra, Phys. Rev. B 81, 104431 (2010)

  16. [24]

    Vidal, Phys

    G. Vidal, Phys. Rev. Lett. 99, 220405 (2007)

  17. [25]

    Verstraete and J

    F. Verstraete and J. I. Cirac, Phys. Rev. Lett.104, 190405 (2010)

  18. [26]

    Haegeman, J

    J. Haegeman, J. I. Cirac, T. J. Osborne, and F. Verstraete, Phys. Rev. B 88, 85118 (2013)

  19. [27]

    Jennings, C

    D. Jennings, C. Brockt, J. Haegeman, T. J. Osborne, and F. Ver- straete, New J. Phys. 17, 063039 (2015)

  20. [28]

    Ganahl and G

    M. Ganahl and G. Vidal, Phys. Rev. B 98, 195105 (2018), arXiv:1801.02219

  21. [29]

    T. J. Osborne (2019), arXiv:1901.06124

  22. [30]

    Tilloy and J

    A. Tilloy and J. I. Cirac, Phys. Rev. X 9, 021040 (2019)

  23. [31]

    De las Cuevas, N

    G. De las Cuevas, N. Schuch, D. Perez-Garcia, and J. I. Cirac, Phys. Rev. B 98, 174303 (2018)

  24. [32]

    Verstraete, J

    F. Verstraete, J. I. Cirac, J. I. Latorre, E. Rico, and M. M. Wolf, Phys. Rev. Lett. 94, 140601 (2005)

  25. [33]

    Schuch, D

    N. Schuch, D. Perez-Garcia, and J. I. Cirac, Ann. Phys. 325, 2153 (2010)

  26. [34]

    Lindblad, Commun

    G. Lindblad, Commun. Math. Phys. 48, 119 (1976)

  27. [35]

    M. M. Wolf (2011), URL https://www-m5.ma.tum. de/foswiki/pub/M5/Allgemeines/MichaelWolf/ QChannelLecture.pdf

  28. [36]

    De las Cuevas and T

    G. De las Cuevas and T. Netzer (2019), arXiv:1907.03664

  29. [37]

    M. M. Wolf and J. I. Cirac, Commun. Math. Phys. 279, 147 (2008)

  30. [38]

    A. S. Kholevo, Theory Probab. Appl. 31, 493 (1987)

  31. [39]

    L. V . Denisov, Theory Probab. Appl.33, 392 (1989)

  32. [40]

    De las Cuevas, J

    G. De las Cuevas, J. I. Cirac, N. Schuch, and D. Perez-Garcia, J. Math. Phys. 58, 121901 (2017)

  33. [41]

    Movassagh and P

    R. Movassagh and P. W. Shor, PNAS 113, 13278 (2016), arXiv:1408.1657

  34. [42]

    Affleck, T

    I. Affleck, T. Kennedy, E. H. Lieb, and H. Tasaki, Comm. Math. Phys. 115, 477 (1988)

  35. [43]

    Haegeman, C

    J. Haegeman, C. Lubich, I. Oseledets, B. Vandereycken, and F. Verstraete, Phys. Rev. B94, 165116 (2016)

  36. [44]

    J. I. Cirac, D. Perez-Garcia, N. Schuch, and F. Verstraete, J. Stat. Mech. p. 083105 (2017)

  37. [45]

    Appendix A: Proof of Proposition 1 In this Appendix we prove Proposition 1

    This is in fact a Matrix Product Vector [ 32], but in this paper we will ignore this distinction and denote unnormalized Matrix Product Vectors as MPS. Appendix A: Proof of Proposition 1 In this Appendix we prove Proposition 1. We start by restat- ing it. Proposition (Proposit...

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.