pith. sign in

arxiv: 2606.09988 · v1 · pith:5CKBRHEQnew · submitted 2026-06-08 · 🪐 quant-ph · cond-mat.dis-nn· cond-mat.str-el· physics.comp-ph

Absence of poor local minima in matrix product states

Hao-Kai Zhang , Chenghong Zhu , Shuo Liu , Shi-Xin Zhang , Tao Xiang This is my paper

Pith reviewed 2026-06-27 16:10 UTC · model grok-4.3

classification 🪐 quant-ph cond-mat.dis-nncond-mat.str-elphysics.comp-ph
keywords matrix product statesenergy landscapelocal minimagauge freedomtrainabilityvariational quantum algorithmssequential circuitsDMRG
0
0 comments X

The pith

Matrix product states have energy landscapes free of poor local minima due to gauge freedom overparametrization.

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

This paper establishes that the energy landscapes of matrix product states contain no poor local minima, resolving why MPS are trainable in practice while general quantum circuits are not. It proves that the gauge freedom in MPS leads to an effective local overparametrization, rendering the distribution of local minima invariant under shifts of the orthogonality center. This invariance causes local minima to concentrate near the global minimum, analogous to overparametrized neural networks. Numerical experiments show that optimization of sequential circuits for MPS converges to near-optimal solutions even for random Hamiltonians, unlike brickwork circuits.

Core claim

The energy landscapes of MPS are free from poor local minima under the same setting where brickwork circuits are not. The local minimum distribution is invariant under moves of the orthogonality center. This invariance arises because the gauge freedom of MPS creates an effective local overparametrization that causes local minima to concentrate near the global minimum.

What carries the argument

Gauge freedom of MPS, which renders the local minimum distribution invariant under orthogonality center moves and creates effective local overparametrization.

If this is right

  • Optimization of sequential circuits for MPS converges to near-optimal solutions even for random Hamiltonians.
  • The local minimum distribution of the MPS energy landscape is invariant under moves of the orthogonality center.
  • This resolves the apparent paradox between trainability issues in quantum circuits and the success of DMRG calculations.
  • Effective local overparametrization determines trainability in variational quantum methods.

Where Pith is reading between the lines

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

  • Tensor network states with similar gauge freedoms may inherit the same absence of poor local minima.
  • Adding local parameter redundancy to other quantum circuit ansatze could improve their optimization landscapes.
  • The overparametrization mechanism may generalize to explain trainability in broader classes of variational quantum models.

Load-bearing premise

The gauge freedom of MPS creates an effective local overparametrization that causes local minima to concentrate near the global minimum, analogous to overparametrized classical neural networks.

What would settle it

An explicit Hamiltonian and initialization where MPS optimization converges to a local minimum with energy significantly above the ground state would falsify the claim.

Figures

Figures reproduced from arXiv: 2606.09988 by Chenghong Zhu, Hao-Kai Zhang, Shi-Xin Zhang, Shuo Liu, Tao Xiang.

Figure 1
Figure 1. Figure 1: FIG. 1. Energy landscape schematics in the absence (a) or [PITH_FULL_IMAGE:figures/full_fig_p002_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Local minimum correspondence. (a) and (c) depict [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: FIG. 3. The numerical training curves of (a) sequential cir [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗
read the original abstract

Quantum circuits suffer from severe trainability issues: even shallow circuits are swamped with poor local minima. Yet matrix product states (MPS), which can be prepared by sequential circuits, are remarkably trainable in practice -- as demonstrated by decades of successful density matrix renormalization group calculations. In this work, we resolve this apparent paradox by proving that the energy landscapes of MPS are free from poor local minima, under the same setting where brickwork circuits are not. The key insight is that the gauge freedom of MPS creates an effective local overparametrization that causes local minima to concentrate near the global minimum, analogous to overparametrized classical neural networks. We rigorously prove that the local minimum distribution of the MPS energy landscape is invariant under moves of the orthogonality center. Numerical experiments further confirm that the optimization of sequential circuits converges to near-optimal solutions even for random Hamiltonians, in stark contrast to brickwork circuits. Our findings highlight the pivotal role of effective local overparametrization in determining trainability, providing a valuable guide for overcoming the trainability bottleneck of variational quantum algorithms.

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 paper claims that matrix product state (MPS) energy landscapes lack poor local minima, resolving the contrast with brickwork quantum circuits. It provides a rigorous proof that the distribution of local minima is invariant under moves of the orthogonality center due to gauge freedom, which is argued to induce effective local overparametrization analogous to overparametrized neural networks. This invariance is said to cause local minima to concentrate near the global minimum. Numerical experiments on random Hamiltonians show that sequential-circuit optimization converges to near-optimal solutions, unlike brickwork circuits.

Significance. If the link between invariance and absence of poor minima holds, the result would explain the empirical success of DMRG and identify effective overparametrization as a key factor in trainability of variational methods. The invariance theorem is a concrete technical contribution, and the numerical contrast with brickwork circuits is informative for VQA design.

major comments (2)
  1. [Abstract / main theorem] Abstract and main theorem: the rigorous result establishes invariance of the local-minimum distribution under orthogonality-center moves, but does not derive that this invariance implies absence of poor local minima or their concentration near the global minimum. The manuscript must supply an explicit argument (beyond the neural-network analogy) showing why gauge-invariant minima cannot remain poor.
  2. [Overparametrization argument] The overparametrization claim: the gauge freedom is said to create 'effective local overparametrization' that forces minima near the global minimum, yet no precise mapping or bound is given that converts the invariance into a concentration statement; this step is load-bearing for the title claim.
minor comments (2)
  1. [Theorem statement] Clarify the precise setting (bond dimension, Hamiltonian class, orthogonality-center definition) under which the invariance theorem holds, and state any assumptions explicitly.
  2. [Numerical experiments] In the numerical section, report the number of random Hamiltonian instances, the precise optimization algorithm and hyperparameters, and quantitative metrics (e.g., energy error relative to exact ground state) to allow direct comparison with brickwork results.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their careful reading and constructive feedback on our manuscript. The comments correctly identify that the link between the invariance theorem and the absence of poor local minima requires a more explicit treatment beyond the analogy. We will revise the manuscript accordingly.

read point-by-point responses

  1. Referee: [Abstract / main theorem] Abstract and main theorem: the rigorous result establishes invariance of the local-minimum distribution under orthogonality-center moves, but does not derive that this invariance implies absence of poor local minima or their concentration near the global minimum. The manuscript must supply an explicit argument (beyond the neural-network analogy) showing why gauge-invariant minima cannot remain poor.

    Authors: We agree that the manuscript would be strengthened by an explicit derivation connecting the invariance result to the absence of poor minima. In the revised version we will add a dedicated subsection after the invariance theorem that directly shows why gauge-invariant local minima cannot remain poor: any candidate poor minimum can be relocated via an orthogonality-center move to a site where the gauge freedom allows a descent direction that reduces the energy below the purported minimum value, contradicting the assumption that it is a local minimum unless its energy equals the global minimum. This argument will be self-contained and will not rely solely on the neural-network analogy. revision: yes

  2. Referee: [Overparametrization argument] The overparametrization claim: the gauge freedom is said to create 'effective local overparametrization' that forces minima near the global minimum, yet no precise mapping or bound is given that converts the invariance into a concentration statement; this step is load-bearing for the title claim.

    Authors: We acknowledge that a precise mapping from invariance to concentration is currently implicit. The revision will introduce a formal definition of effective local overparametrization for MPS (quantifying the redundant degrees of freedom per site after fixing the orthogonality center) and will prove a concentration bound: the invariance implies that the set of local minima is closed under gauge transformations, which in turn forces the energy values at local minima to lie within an additive factor proportional to the bond dimension of the global minimum energy. This bound will be stated as a new theorem supporting the title claim. revision: yes

Circularity Check

0 steps flagged

No circularity: mathematical invariance proof is independent of the overparametrization interpretation

full rationale

The central result is a claimed rigorous proof that the MPS energy landscape's local minimum distribution is invariant under orthogonality center moves. This is presented as a first-principles mathematical statement, not derived from a fit, self-definition, or prior self-citation chain. The further interpretive step linking gauge freedom to concentration near the global minimum (via neural-network analogy) is supported by numerics rather than reducing the proof itself to an input by construction. No load-bearing self-citations, ansatz smuggling, or renaming of known results appear in the derivation chain as described.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review prevents full enumeration; the result appears to rest on standard mathematical properties of tensor networks and optimization landscapes with no free parameters or invented entities identified.

pith-pipeline@v0.9.1-grok · 5734 in / 1057 out tokens · 30496 ms · 2026-06-27T16:10:29.316681+00:00 · methodology

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

46 extracted references · 36 canonical work pages

  1. [1]

    Choromanska , author M

    author author A. Choromanska , author M. Henaff , author M. Mathieu , author G. B. \ Arous , \ and\ author Y. LeCun ,\ title The Loss Surfaces of Multilayer Networks ,\ https://proceedings.mlr.press/v38/choromanska15 journal journal Journal of Machine Learning Research \ volume 38 ,\ pages 192 ( year 2015 ) NoStop

    2015

  2. [2]

    Noisy intermediate-scale quantum algorithms,

    author author K. Bharti , author A. Cervera-Lierta , author T. H. \ Kyaw , author T. Haug , author S. Alperin-Lea , author A. Anand , author M. Degroote , author H. Heimonen , author J. S. \ Kottmann , author T. Menke , author W.-K. \ Mok , author S. Sim , author L.-C. \ Kwek , \ and\ author A. Aspuru-Guzik ,\ title Noisy intermediate-scale quantum algori...

    work page doi:10.1103/revmodphys.94.015004 2022

  3. [3]

    Nature Reviews Physics , author =

    author author M. Cerezo , author A. Arrasmith , author R. Babbush , author S. C. \ Benjamin , author S. Endo , author K. Fujii , author J. R. \ McClean , author K. Mitarai , author X. Yuan , author L. Cincio , \ and\ author P. J. \ Coles ,\ title Variational quantum algorithms ,\ 10.1038/s42254-021-00348-9 journal journal Nature Reviews Physics \ volume 3...

    work page doi:10.1038/s42254-021-00348-9 2021

  4. [4]

    author author J. R. \ McClean , author S. Boixo , author V. N. \ Smelyanskiy , author R. Babbush , \ and\ author H. Neven ,\ title Barren plateaus in quantum neural network training landscapes ,\ 10.1038/s41467-018-07090-4 journal journal Nature Communications \ volume 9 ,\ pages 1 ( year 2018 ) NoStop

    work page doi:10.1038/s41467-018-07090-4 2018

  5. [5]

    Cerezo , author A

    author author M. Cerezo , author A. Sone , author T. Volkoff , author L. Cincio , \ and\ author P. J. \ Coles ,\ title Cost function dependent barren plateaus in shallow parametrized quantum circuits ,\ 10.1038/s41467-021-21728-w journal journal Nature Communications \ volume 12 ,\ pages 1791 ( year 2021 b ) NoStop

    work page doi:10.1038/s41467-021-21728-w 2021

  6. [6]

    \ Zhang , author S

    author author H.-K. \ Zhang , author S. Liu , \ and\ author S.-X. \ Zhang ,\ title Absence of Barren Plateaus in Finite Local-Depth Circuits with Long-Range Entanglement ,\ 10.1103/PhysRevLett.132.150603 journal journal Physical Review Letters \ volume 132 ,\ pages 150603 ( year 2024 a ) NoStop

    work page doi:10.1103/physrevlett.132.150603 2024

  7. [7]

    and Cincio, Lukasz and McClean, Jarrod R

    author author M. Larocca , author S. Thanasilp , author S. Wang , author K. Sharma , author J. Biamonte , author P. J. \ Coles , author L. Cincio , author J. R. \ McClean , author Z. Holmes , \ and\ author M. Cerezo ,\ title Barren plateaus in variational quantum computing ,\ 10.1038/s42254-025-00813-9 journal journal Nature Reviews Physics \ volume 7 ,\ ...

    work page doi:10.1038/s42254-025-00813-9 2025

  8. [8]

    Bittel \ and\ author M

    author author L. Bittel \ and\ author M. Kliesch ,\ title Training Variational Quantum Algorithms Is NP-Hard ,\ 10.1103/PhysRevLett.127.120502 journal journal Physical Review Letters \ volume 127 ,\ pages 120502 ( year 2021 ) NoStop

    work page doi:10.1103/physrevlett.127.120502 2021

  9. [9]

    You \ and\ author X

    author author X. You \ and\ author X. Wu ,\ Exponentially Many Local Minima in Quantum Neural Networks ,\ in\ http://arxiv.org/abs/2110.02479 booktitle Proceedings of Machine Learning Research ,\ Vol.\ volume 139 \ ( year 2021 )\ pp.\ pages 12144--12155 NoStop

    arXiv 2021

  10. [10]

    author author E. R. \ Anschuetz \ and\ author B. T. \ Kiani ,\ title Quantum variational algorithms are swamped with traps ,\ https://www.nature.com/articles/s41467-022-35364-5 journal journal Nature Communications \ volume 13 ( year 2022 ) NoStop

    2022

  11. [11]

    author author E. R. \ Anschuetz ,\ A Unified Theory of Quantum Neural Network Loss Landscapes ,\ in\ http://arxiv.org/abs/2408.11901 booktitle International Conference on Learning Representations ,\ series and number number 1 \ ( year 2025 )\ pp.\ pages 1--60 NoStop

    arXiv 2025

  12. [12]

    Quantum , author =

    author author A. Arrasmith , author M. Cerezo , author P. Czarnik , author L. Cincio , \ and\ author P. J. \ Coles ,\ title Effect of barren plateaus on gradient-free optimization ,\ 10.22331/q-2021-10-05-558 journal journal Quantum \ volume 5 ,\ pages 558 ( year 2021 ) NoStop

    work page doi:10.22331/q-2021-10-05-558 2021

  13. [13]

    Arrasmith , author Z

    author author A. Arrasmith , author Z. Holmes , author M. Cerezo , \ and\ author P. J. \ Coles ,\ title Equivalence of quantum barren plateaus to cost concentration and narrow gorges ,\ 10.1088/2058-9565/ac7d06 journal journal Quantum Science and Technology \ volume 7 ,\ pages 045015 ( year 2022 ) NoStop

    work page doi:10.1088/2058-9565/ac7d06 2058

  14. [14]

    Liu , author L.-W

    author author Z. Liu , author L.-W. \ Yu , author L.-M. \ Duan , \ and\ author D.-L. \ Deng ,\ title Presence and Absence of Barren Plateaus in Tensor-Network Based Machine Learning ,\ 10.1103/PhysRevLett.129.270501 journal journal Physical Review Letters \ volume 129 ,\ pages 270501 ( year 2022 ) NoStop

    work page doi:10.1103/physrevlett.129.270501 2022

  15. [15]

    Holmes , author K

    author author Z. Holmes , author K. Sharma , author M. Cerezo , \ and\ author P. J. \ Coles ,\ title Connecting Ansatz Expressibility to Gradient Magnitudes and Barren Plateaus ,\ 10.1103/PRXQuantum.3.010313 journal journal PRX Quantum \ volume 3 ,\ pages 010313 ( year 2022 ) NoStop

    work page doi:10.1103/prxquantum.3.010313 2022

  16. [16]

    \ Zhang , author C

    author author H.-K. \ Zhang , author C. Zhu , author M. Jing , \ and\ author X. Wang ,\ title Statistical Analysis of Quantum State Learning Process in Quantum Neural Networks ,\ http://arxiv.org/abs/2309.14980 journal journal Advances in Neural Information Processing Systems \ ( year 2023 a ) NoStop

    arXiv 2023

  17. [17]

    Barthel \ and\ author Q

    author author T. Barthel \ and\ author Q. Miao ,\ title Absence of Barren Plateaus and Scaling of Gradients in the Energy Optimization of Isometric Tensor Network States ,\ 10.1007/s00220-024-05217-x journal journal Communications in Mathematical Physics \ volume 406 ,\ pages 86 ( year 2025 ) NoStop

    work page doi:10.1007/s00220-024-05217-x 2025

  18. [18]

    Miao \ and\ author T

    author author Q. Miao \ and\ author T. Barthel ,\ title Isometric tensor network optimization for extensive Hamiltonians is free of barren plateaus ,\ 10.1103/PhysRevA.109.L050402 journal journal Physical Review A \ volume 109 ,\ pages L050402 ( year 2024 ) NoStop

    work page doi:10.1103/physreva.109.l050402 2024

  19. [19]

    Liu , author S.-X

    author author S. Liu , author S.-X. \ Zhang , author S.-K. \ Jian , \ and\ author H. Yao ,\ title Training variational quantum algorithms with random gate activation ,\ 10.1103/PhysRevResearch.5.L032040 journal journal Physical Review Research \ volume 5 ,\ pages L032040 ( year 2023 ) NoStop

    work page doi:10.1103/physrevresearch.5.l032040 2023

  20. [20]

    Liu , author G

    author author X. Liu , author G. Liu , author H.-K. \ Zhang , author J. Huang , \ and\ author X. Wang ,\ title Mitigating Barren Plateaus of Variational Quantum Eigensolvers ,\ 10.1109/TQE.2024.3383050 journal journal IEEE Transactions on Quantum Engineering \ volume 5 ,\ pages 1 ( year 2024 ) NoStop

    work page doi:10.1109/tqe.2024.3383050 2024

  21. [21]

    \ Zhang , author C

    author author H.-K. \ Zhang , author C. Zhu , author G. Liu , \ and\ author X. Wang ,\ title Exponential Hardness of Optimization from the Locality in Quantum Neural Networks ,\ 10.1609/aaai.v38i15.29614 journal journal Proceedings of the AAAI Conference on Artificial Intelligence \ volume 38 ,\ pages 16741 ( year 2024 b ) NoStop

    work page doi:10.1609/aaai.v38i15.29614 2024

  22. [22]

    \ Zhang , author C

    author author H.-K. \ Zhang , author C. Zhu , \ and\ author X. Wang ,\ title Predicting quantum learnability from landscape fluctuation ,\ http://arxiv.org/abs/2406.11805 journal journal arXiv:2406.11805 \ ( year 2024 c ) NoStop

    arXiv 2024

  23. [23]

    Cerezo , author M

    author author M. Cerezo , author M. Larocca , author D. Garc \' i a-Mart \' i n , author N. L. \ Diaz , author P. Braccia , author E. Fontana , author M. S. \ Rudolph , author P. Bermejo , author A. Ijaz , author S. Thanasilp , author E. R. \ Anschuetz , \ and\ author Z. Holmes ,\ title Does provable absence of barren plateaus imply classical simulability...

    work page doi:10.1038/s41467-025-63099-6 2025

  24. [24]

    author author S. R. \ White ,\ title Density matrix formulation for quantum renormalization groups ,\ 10.1103/PhysRevLett.69.2863 journal journal Physical Review Letters \ volume 69 ,\ pages 2863 ( year 1992 ) NoStop

    work page doi:10.1103/physrevlett.69.2863 1992

  25. [25]

    author author U. Schollw \" o ck ,\ title The density-matrix renormalization group in the age of matrix product states ,\ 10.1016/j.aop.2010.09.012 journal journal Annals of Physics \ volume 326 ,\ pages 96 ( year 2011 ) NoStop

    work page doi:10.1016/j.aop.2010.09.012 2010

  26. [26]

    author author R. Or \' u s ,\ title A practical introduction to tensor networks: Matrix product states and projected entangled pair states ,\ 10.1016/j.aop.2014.06.013 journal journal Annals of Physics \ volume 349 ,\ pages 117 ( year 2014 ) NoStop

    work page doi:10.1016/j.aop.2014.06.013 2014

  27. [27]

    Or \' u s ,\ title Tensor networks for complex quantum systems ,\ 10.1038/s42254-019-0086-7 journal journal Nature Reviews Physics \ volume 1 ,\ pages 538 ( year 2019 ) NoStop

    author author R. Or \' u s ,\ title Tensor networks for complex quantum systems ,\ 10.1038/s42254-019-0086-7 journal journal Nature Reviews Physics \ volume 1 ,\ pages 538 ( year 2019 ) NoStop

    work page doi:10.1038/s42254-019-0086-7 2019

  28. [28]

    author author J. I. \ Cirac , author D. P \' e rez-Garc \' i a , author N. Schuch , \ and\ author F. Verstraete ,\ title Matrix product states and projected entangled pair states: Concepts, symmetries, theorems ,\ 10.1103/RevModPhys.93.045003 journal journal Reviews of Modern Physics \ volume 93 ,\ pages 045003 ( year 2021 ) NoStop

    work page doi:10.1103/revmodphys.93.045003 2021

  29. [29]

    Xiang ,\ Density Matrix and Tensor Network Renormalization \ ( publisher Cambridge University Press ,\ year 2023 ) NoStop

    author author T. Xiang ,\ Density Matrix and Tensor Network Renormalization \ ( publisher Cambridge University Press ,\ year 2023 ) NoStop

    2023

  30. [30]

    Ignacio and Osborne, Tobias J

    author author J. Haegeman , author J. I. \ Cirac , author T. J. \ Osborne , author I. Pi z orn , author H. Verschelde , \ and\ author F. Verstraete ,\ title Time-Dependent Variational Principle for Quantum Lattices ,\ 10.1103/PhysRevLett.107.070601 journal journal Physical Review Letters \ volume 107 ,\ pages 070601 ( year 2011 ) NoStop

    work page doi:10.1103/physrevlett.107.070601 2011

  31. [31]

    Haegeman , author M

    author author J. Haegeman , author M. Marien , author T. J. \ Osborne , \ and\ author F. Verstraete ,\ title Geometry of matrix product states: Metric, parallel transport, and curvature ,\ https://pubs.aip.org/jmp/article/55/2/021902/232472/Geometry-of-matrix-product-states-Metric-parallel journal journal Journal of Mathematical Physics \ volume 55 ( year...

    2014

  32. [32]

    doi:10.1103/PhysRevB.94.165116 , url =

    author author J. Haegeman , author C. Lubich , author I. Oseledets , author B. Vandereycken , \ and\ author F. Verstraete ,\ title Unifying time evolution and optimization with matrix product states ,\ 10.1103/PhysRevB.94.165116 journal journal Physical Review B \ volume 94 ,\ pages 165116 ( year 2016 ) NoStop

    work page doi:10.1103/physrevb.94.165116 2016

  33. [33]

    Hauru , author M

    author author M. Hauru , author M. Van Damme , \ and\ author J. Haegeman ,\ title Riemannian optimization of isometric tensor networks ,\ 10.21468/SciPostPhys.10.2.040 journal journal SciPost Physics \ volume 10 ,\ pages 040 ( year 2021 ) NoStop

    work page doi:10.21468/scipostphys.10.2.040 2021

  34. [34]

    author author M. B. \ Hastings ,\ title An area law for one-dimensional quantum systems ,\ 10.1088/1742-5468/2007/08/P08024 journal journal Journal of Statistical Mechanics: Theory and Experiment \ volume 2007 ,\ pages P08024 ( year 2007 ) NoStop

    work page doi:10.1088/1742-5468/2007/08/p08024 2007

  35. [35]

    Eisert , author M

    author author J. Eisert , author M. Cramer , \ and\ author M. B. \ Plenio ,\ title Colloquium : Area laws for the entanglement entropy ,\ 10.1103/RevModPhys.82.277 journal journal Reviews of Modern Physics \ volume 82 ,\ pages 277 ( year 2010 ) NoStop

    work page doi:10.1103/revmodphys.82.277 2010

  36. [36]

    @noop See the Supplemental Material for preliminaries on matrix product states and the Weingarten calculus, rigorous theorem proofs, and additional numerical results. Stop

  37. [37]

    Stokes, J

    author author J. Stokes , author J. Izaac , author N. Killoran , \ and\ author G. Carleo ,\ title Quantum Natural Gradient ,\ 10.22331/q-2020-05-25-269 journal journal Quantum \ volume 4 ,\ pages 269 ( year 2020 ) NoStop

    work page doi:10.22331/q-2020-05-25-269 2020

  38. [38]

    Larocca , author N

    author author M. Larocca , author N. Ju , author D. Garc \' i a-Mart \' i n , author P. J. \ Coles , \ and\ author M. Cerezo ,\ title Theory of overparametrization in quantum neural networks ,\ 10.1038/s43588-023-00467-6 journal journal Nature Computational Science \ volume 3 ,\ pages 542 ( year 2023 ) NoStop

    work page doi:10.1038/s43588-023-00467-6 2023

  39. [39]

    Collins \ and\ author P

    author author B. Collins \ and\ author P. \' S niady ,\ title Integration with Respect to the Haar Measure on Unitary, Orthogonal and Symplectic Group ,\ 10.1007/s00220-006-1554-3 journal journal Communications in Mathematical Physics \ volume 264 ,\ pages 773 ( year 2006 ) NoStop

    work page doi:10.1007/s00220-006-1554-3 2006

  40. [40]

    Zhou \ and\ author A

    author author T. Zhou \ and\ author A. Nahum ,\ title Emergent statistical mechanics of entanglement in random unitary circuits ,\ 10.1103/PhysRevB.99.174205 journal journal Physical Review B \ volume 99 ,\ pages 174205 ( year 2019 ) NoStop

    work page doi:10.1103/physrevb.99.174205 2019

  41. [41]

    author author M. P. \ Zaletel \ and\ author F. Pollmann ,\ title Isometric Tensor Network States in Two Dimensions ,\ 10.1103/PhysRevLett.124.037201 journal journal Physical Review Letters \ volume 124 ,\ pages 037201 ( year 2020 ) NoStop

    work page doi:10.1103/physrevlett.124.037201 2020

  42. [42]

    \ Wei , author D

    author author Z.-Y. \ Wei , author D. Malz , \ and\ author J. I. \ Cirac ,\ title Sequential Generation of Projected Entangled-Pair States ,\ 10.1103/PhysRevLett.128.010607 journal journal Physical Review Letters \ volume 128 ,\ pages 010607 ( year 2022 ) NoStop

    work page doi:10.1103/physrevlett.128.010607 2022

  43. [43]

    author author B. Kutschan ,\ title Tangent cones to tensor train varieties ,\ 10.1016/j.laa.2018.01.012 journal journal Linear Algebra and its Applications \ volume 544 ,\ pages 370 ( year 2018 ) NoStop

    work page doi:10.1016/j.laa.2018.01.012 2018

  44. [44]

    2023 , pages =

    author author S.-X. \ Zhang , author J. Allcock , author Z.-Q. \ Wan , author S. Liu , author J. Sun , author H. Yu , author X.-H. \ Yang , author J. Qiu , author Z. Ye , author Y.-Q. \ Chen , author C.-K. \ Lee , author Y.-C. \ Zheng , author S.-K. \ Jian , author H. Yao , author C.-Y. \ Hsieh , \ and\ author S. Zhang ,\ title TensorCircuit: a Quantum So...

    work page doi:10.22331/q-2023-02-02-912 2023

  45. [45]

    \ Zhang , author Y.-Q

    author author S.-X. \ Zhang , author Y.-Q. \ Chen , author W. Li , author J. Sun , author W.-G. \ Ma , author P.-L. \ Zheng , author Y.-X. \ Huang , author Q.-X. \ Wang , author H. Yu , author Z. Li , author X. Huang , author Z.-L. \ Li , author Z.-Q. \ Wan , author S. Liu , author J. Qiu , author J. Miao , author Z. Song , author Y. Yan , author K. Tsuok...

    arXiv 2026

  46. [46]

    Drury \ and\ author P

    author author B. Drury \ and\ author P. Love ,\ title Constructive quantum Shannon decomposition from Cartan involutions ,\ 10.1088/1751-8113/41/39/395305 journal journal Journal of Physics A: Mathematical and Theoretical \ volume 41 ,\ pages 395305 ( year 2008 ) NoStop

    work page doi:10.1088/1751-8113/41/39/395305 2008