Frustration graph formalism for qudit observables

B{\l}a\.zej Kuzaka; Owidiusz Makuta; Remigiusz Augusiak

arxiv: 2503.22400 · v3 · pith:TYOFUMTInew · submitted 2025-03-28 · 🪐 quant-ph

Frustration graph formalism for qudit observables

Owidiusz Makuta , B{\l}a\.zej Kuzaka , Remigiusz Augusiak This is my paper

Pith reviewed 2026-05-22 22:35 UTC · model grok-4.3

classification 🪐 quant-ph

keywords qudit observablesfrustration graphgeneralized Pauli matricescommutation relationsstabilizer subspacesentanglement measureprime dimensionunitary transformation

0 comments

The pith

For groups of prime-d qudit observables that commute up to root-of-unity factors, a single unitary maps them all to generalized Pauli matrices tensored with commuting operators.

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

The paper examines groups of d-outcome observables for prime d, represented as non-Hermitian unitaries with d-th root eigenvalues. These observables satisfy commutation relations up to scalar factors that are also roots of unity. By encoding these relations in a frustration graph, the authors prove that any such group can be simultaneously transformed by one unitary operator into a product of generalized Pauli matrices and some mutually commuting ancillary operators. This transformation yields upper bounds on the sum of squared absolute values of the observables and on the sum of their expectation values. The bounds are then applied to calculate the generalized geometric measure of entanglement in qudit stabilizer subspaces.

Core claim

By representing the commutation relations of these observables via a frustration graph, we show that for such a group, there exists a single unitary transforming them into a tensor product of generalized Pauli matrices and some ancillary mutually commuting operators. Building on this result, we derive upper bounds on the sum of the squares of the absolute values and the sum of the expected values of the observables forming a group. We finally utilize these bounds to compute the generalized geometric measure of entanglement for qudit stabilizer subspaces.

What carries the argument

The frustration graph that encodes the commutation relations (up to roots of unity) among the observables, allowing reduction to a standard form via a single unitary.

Load-bearing premise

The observables must be non-Hermitian unitaries with eigenvalues that are d-th roots of unity for prime d, and they commute only up to multiplication by such a root of unity.

What would settle it

Finding a specific set of such observables for prime d where no single unitary exists that maps the entire group to generalized Paulis plus commuting operators would falsify the claim.

read the original abstract

The incompatibility of measurements is the key feature of quantum theory that distinguishes it from the classical description of nature. Here, we consider groups of d-outcome quantum observables with prime d represented by non-Hermitian unitary operators whose eigenvalues are d'th roots of unity. We additionally assume that these observables mutually commute up to a scalar factor being one of the d'th roots of unity. By representing commutation relations of these observables via a frustration graph, we show that for such a group, there exists a single unitary transforming them into a tensor product of generalized Pauli matrices and some ancillary mutually commuting operators. Building on this result, we derive upper bounds on the sum of the squares of the absolute values and the sum of the expected values of the observables forming a group. We finally utilize these bounds to compute the generalized geometric measure of entanglement for qudit stabilizer subspaces.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper repackages the uniqueness of Heisenberg-group irreps over finite fields into a frustration-graph construction for prime-d qudit observables and extracts concrete bounds for entanglement calculations.

read the letter

The central move is to encode the root-of-unity phase factors in the commutation relations as a frustration graph, then show that any such closed group admits a single unitary that puts every operator into tensor-product generalized Pauli form plus some mutually commuting ancillas. From there the authors derive upper bounds on the sum of squared moduli and on the sum of expectation values, and plug those bounds into the generalized geometric measure of entanglement for stabilizer subspaces in prime dimension. That last step is the part that could actually be used in calculations. The graph step makes the cocycle visible and the reduction step avoids having to handle each set of observables separately, so the formalism is tidy for this restricted class. The underlying fact that all irreps are unitarily equivalent is standard representation theory of the finite Heisenberg group, so the paper is not claiming a new theorem there; its contribution is the packaging and the downstream bounds. No circularity appears in the logic, and the assumptions on the observables match the usual setup for these groups. The bounds themselves are not claimed to be tight, but they are explicit enough to be checked against small cases. This is aimed at people already working on qudit stabilizer codes and entanglement monotones in prime dimensions. A reader who needs a systematic way to bound incompatibility relations in that setting will find the construction helpful. The work is coherent on its own terms and the math checks out against known facts, so it deserves a serious referee even though the core equivalence is not novel.

Referee Report

2 major / 2 minor

Summary. The manuscript develops a frustration-graph formalism for finite groups of d-outcome observables (d prime) realized by non-Hermitian unitaries whose eigenvalues are d-th roots of unity and that commute up to d-th-root phases. It asserts that any such group admits a single unitary conjugating the entire set into a tensor-product form consisting of generalized Pauli operators together with ancillary mutually commuting operators. From this structural result the authors derive two families of upper bounds—one on the sum of squares of absolute values and one on the sum of expectation values—and apply the bounds to obtain the generalized geometric measure of entanglement for qudit stabilizer subspaces.

Significance. If the central structural claim holds, the work supplies a concrete graph-theoretic device for reducing the analysis of qudit incompatibility to the representation theory of the Heisenberg group over F_d. The resulting bounds are parameter-free once the graph is given and therefore furnish falsifiable, computable limits on entanglement witnesses for stabilizer subspaces. The approach is a direct qudit extension of existing qubit frustration-graph techniques and, when accompanied by explicit code or machine-checked proofs, would constitute a reusable tool for higher-dimensional stabilizer theory.

major comments (2)

[§3, Theorem 1] §3, Theorem 1: the existence of the conjugating unitary is stated to follow from the frustration-graph encoding of the phase cocycle together with uniqueness of the irreducible projective representation. The proof sketch does not explicitly verify that the dimension of the Hilbert space on which the representation is realized is a power of d; without this step the appeal to uniqueness of the Weyl representation is incomplete.
[§4.2, Eq. (18)] §4.2, Eq. (18): the bound on ∑|⟨O_i⟩|² is obtained by summing the squared norms after conjugation. The derivation assumes that the ancillary commuting operators contribute zero to the sum; this holds only if those operators are traceless on the support of the state under consideration. The manuscript does not state the support condition explicitly.

minor comments (2)

Notation: the symbol G is used both for the abstract group and for its frustration graph; a distinct symbol for the graph would improve readability.
Figure 2: the edge labels indicating the phase factors are too small to read at standard print size; enlarging the labels or adding a legend would help.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful reading and constructive comments on the manuscript. We address each major comment below.

read point-by-point responses

Referee: [§3, Theorem 1] the existence of the conjugating unitary is stated to follow from the frustration-graph encoding of the phase cocycle together with uniqueness of the irreducible projective representation. The proof sketch does not explicitly verify that the dimension of the Hilbert space on which the representation is realized is a power of d; without this step the appeal to uniqueness of the Weyl representation is incomplete.

Authors: We agree that the proof sketch should explicitly verify the Hilbert-space dimension. The observables act on a tensor-product space of n qudits, each of local dimension d, so the total dimension is d^n. We will revise the proof of Theorem 1 in §3 to include this verification before invoking uniqueness of the irreducible projective representation of the Heisenberg group over F_d. revision: yes
Referee: [§4.2, Eq. (18)] the bound on ∑|⟨O_i⟩|² is obtained by summing the squared norms after conjugation. The derivation assumes that the ancillary commuting operators contribute zero to the sum; this holds only if those operators are traceless on the support of the state under consideration. The manuscript does not state the support condition explicitly.

Authors: The referee is correct that the support condition is not stated explicitly. We will revise §4.2 to add the explicit requirement that the bound applies to states supported on the subspace on which the ancillary operators are traceless (equivalently, have vanishing expectation value). revision: yes

Circularity Check

0 steps flagged

No significant circularity

full rationale

The derivation begins from the stated assumptions on the observables (non-Hermitian unitaries with d-th root eigenvalues, d prime, closed under multiplication, commuting up to phases) and encodes their relations in a frustration graph. The existence of the conjugating unitary is then shown to follow from this encoding together with standard facts about projective representations of the Heisenberg group over F_d. The subsequent bounds on sums of squares and expectation values, and the generalized geometric measure of entanglement, are direct consequences of that unitary equivalence. No step reduces a claimed prediction or theorem to a fitted parameter, a self-definition, or a load-bearing self-citation whose content is itself unverified. The central claim is mathematically independent of the final entanglement application and does not rename a known empirical pattern as a new result.

Axiom & Free-Parameter Ledger

0 free parameters · 2 axioms · 1 invented entities

The work rests on standard quantum-mechanical assumptions about unitary observables and introduces the frustration graph as a modeling device; no free parameters or new physical entities are introduced.

axioms (2)

domain assumption Observables are non-Hermitian unitary operators with eigenvalues equal to d-th roots of unity for prime d.
This representation is stated as the starting point for the observables under consideration.
domain assumption The observables commute up to multiplication by a d-th root of unity scalar.
This near-commutation condition defines the groups to which the graph formalism applies.

invented entities (1)

Frustration graph for qudit observables no independent evidence
purpose: To encode the commutation relations among the observables.
The graph is introduced in the paper as the central representational tool.

pith-pipeline@v0.9.0 · 5682 in / 1417 out tokens · 35484 ms · 2026-05-22T22:35:45.239017+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Foundation/AlexanderDuality.lean alexander_duality_circle_linking unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

By representing commutation relations of these observables via a frustration graph, we show that for such a group, there exists a single unitary transforming them into a tensor product of generalized Pauli matrices and some ancillary mutually commuting operators.
IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Theorem 2. ... sum |<A>|^2 ≤ d^(null(γ)+k)/2 = ω̃(G)

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

Entanglement witnesses for stabilizer states and subspaces beyond qubits
quant-ph 2025-08 unverdicted novelty 6.0

Generalizes entanglement witnesses from qubit stabilizer states to multi-qudit versions, showing better noise robustness in some cases.

Reference graph

Works this paper leans on

49 extracted references · 49 canonical work pages · cited by 1 Pith paper · 1 internal anchor

[1]

Notice that {Ti}k i=1 is a generating set of A and the group operation in A is given byAI ⊙ AJ =AI+J , so thatI +J ∈ Zk d

such that • Ad I = 1 for anyI, • for any pair AI,A J ∈ A we have [AI,A J ]• =ω l1 for somel ∈ Zd. Notice that {Ti}k i=1 is a generating set of A and the group operation in A is given byAI ⊙ AJ =AI+J , so thatI +J ∈ Zk d. However, it is important to state that in this work, we are interested in commutation relations of elements A with respect to regular ma...

work page
[2]

Theorem 2

is in fact bounded from above by ˜ω (G). Theorem 2. Let A = ⟨T1,...,T k⟩⊙ be a group as in Deﬁni- tion 1. F or each such A we have ∑ A∈A |⟨A⟩|2 ⩽d(null(γ)+k)/2 = ˜ω (G). (28) The detailed proof can be found in Appendix B. The main idea of the proof is to rewrite the sum of squares in Eq. ( 27) as ∑ A∈A |⟨A⟩|2 = tr [ ∑ A∈A (A ⊗ A†)ρ ⊗ ρ ] , (29) which then...

work page 2019
[3]

M. T. Quintino, T. Vértesi, and N. Brunner, Phys. Rev. Lett. 113, 160402 (2014)

work page 2014
[4]

Mayers and A

D. Mayers and A. Yao, in Proceedings 39th Annual Symposium on F oundations of Computer Science (Cat. No.98CB36280) (1998) pp. 503–509

work page 1998
[5]

Zapatero, T

V . Zapatero, T. van Leent, R. Arnon-Friedman, W.-Z. Liu, Q. Zhang, H. Weinfurter, and M. Curty, npj Quantum Infor- mation 9, 10 (2023)

work page 2023
[6]

D. P . Nadlinger, P . Drmota, B. C. Nichol, G. Araneda, D. Main, R. Srinivas, D. M. Lucas, C. J. Ballance, K. Ivanov, E. Y .-Z. Tan, P . Sekatski, R. L. Urbanke, R. Renner, N. Sangouard, and J.-D. Bancal, Nature 607, 682–686 (2022)

work page 2022
[7]

Šupi ´c and J

I. Šupi ´c and J. Bowles, Quantum 4, 337 (2020)

work page 2020
[8]

Baccari, R

F. Baccari, R. Augusiak, I. Šupi ´c, and A. Acín, Physical Review Letters 125 (2020), 10.1103/physrevlett.125.260507

work page doi:10.1103/physrevlett.125.260507 2020
[9]

Acín and L

A. Acín and L. Masanes, Nature 540, 213 (2016). 8

work page 2016
[10]

Pironio, A

S. Pironio, A. Acín, S. Massar, A. B. de la Giroday, D. N. Mat- sukevich, P . Maunz, S. Olmschenk, D. Hayes, L. Luo, T. A. Manning, and C. Monroe, Nature 464, 1021 (2010)

work page 2010
[11]

Maccone and A

L. Maccone and A. K. Pati, Phys. Rev. Lett. 113, 260401 (2014)

work page 2014
[12]

de Guise, L

H. de Guise, L. Maccone, B. C. Sanders, and N. Shukla, Phys. Rev. A 98, 042121 (2018)

work page 2018
[13]

Y . Xiao, C. Guo, F. Meng, N. Jing, and M.-H. Y ung, Phys. Rev. A 100, 032118 (2019)

work page 2019
[14]

Frustrated spin systems: History of the emergence of a mode rn physics,

H. T. Diep,“Frustrated spin systems: History of the emergence of a mode rn physics,” (2024), arXiv:2411.12826 [cond-mat.stat-mech]

work page arXiv 2024
[15]

M. B. Hastings and R. O’Donnell, in Proceedings of the 54th Annual ACM SIGACT Symposium on Theor y of Computing , STOC 2022 (Association for Computing Machinery, New Y ork, NY , USA, 2022) p. 776–789

work page 2022
[16]

Chapman and S

A. Chapman and S. T. Flammia, Quantum 4, 278 (2020)

work page 2020
[17]

Chapman, S

A. Chapman, S. J. Elman, and R. L. Mann, “A uniﬁed graph-theoretic framework for free-fermion solv ability,” (2023), arXiv:2305.15625 [quant-ph]

work page arXiv 2023
[18]

de Gois, K

C. de Gois, K. Hansenne, and O. Gühne, Phys. Rev. A 107, 062211 (2023)

work page 2023
[19]

Z.-P . Xu, R. Schwonnek, and A. Winter, PRX Quantum 5 (2024), 10.1103/prxquantum.5.020318

work page doi:10.1103/prxquantum.5.020318 2024
[20]

Full classiﬁcation of pauli lie algebras,

G. Aguilar, S. Cichy, J. Eisert, and L. Bit- tel, “Full classiﬁcation of pauli lie algebras,” (2024), arXiv:2408.00081 [quant-ph]

work page arXiv 2024
[21]

Lovasz, IEEE Transactions on Information Theory 25, 1 (1979)

L. Lovasz, IEEE Transactions on Information Theory 25, 1 (1979)

work page 1979
[22]

A graph-theoretic framework for free-parafermion solvab ility,

R. L. Mann, S. J. Elman, D. R. Wood, and A. Chapman, “A graph-theoretic framework for free-parafermion solvab ility,” (2024), arXiv:2408.09684 [quant-ph]

work page arXiv 2024
[23]

Sarkar and T

R. Sarkar and T. J. Y oder,Quantum 8, 1307 (2024)

work page 2024
[24]

Santos, C

R. Santos, C. Jebarathinam, and R. Augusiak, Phys. Rev. A 106, 012431 (2022)

work page 2022
[25]

A. M. Steane, Phys. Rev. Lett. 77, 793 (1996)

work page 1996
[26]

P . J. Nadkarni and S. S. Garani, Phys. Rev. A 103, 042420 (2021)

work page 2021
[27]

Stabilizer Codes and Quantum Error Correction

D. Gottesman, “Stabilizer codes and quantum error correction,” (1997), arXiv:quant-ph/9705052 [quant-ph]

work page internal anchor Pith review Pith/arXiv arXiv 1997
[28]

Tóth and O

G. Tóth and O. Gühne, Physical Review A 72 (2005), 10.1103/physreva.72.022340

work page doi:10.1103/physreva.72.022340 2005
[29]

Tóth and O

G. Tóth and O. Gühne, Phys. Rev. A 72, 022340 (2005)

work page 2005
[30]

Makuta and R

O. Makuta and R. Augusiak, New Journal of Physics 23, 043042 (2021)

work page 2021
[31]

Sen(De) and U

A. Sen(De) and U. Sen, Phys. Rev. A 81, 012308 (2010)

work page 2010
[32]

Seevinck and J

M. Seevinck and J. Ufﬁnk, Phys. Rev. A 65, 012107 (2001)

work page 2001
[33]

SHIMONY ,Annals of the New Y ork Academy of Sciences 755, 675 (1995) https://nyaspubs.onlinelibrary.wiley.com/doi/pdf/10.1111/j.1749-6632.1995.tb3

A. SHIMONY ,Annals of the New Y ork Academy of Sciences 755, 675 (1995) https://nyaspubs.onlinelibrary.wiley.com/doi/pdf/10.1111/j.1749-6632.1995.tb3

work page doi:10.1111/j.1749-6632.1995.tb3 1995
[34]

Barnum and N

H. Barnum and N. Linden, Journal of Physics A: Mathematical and General 34, 6787 (2001)

work page 2001
[35]

Gour and N

G. Gour and N. R. Wallach, Phys. Rev. A 76, 042309 (2007)

work page 2007
[36]

Branciard, H

C. Branciard, H. Zhu, L. Chen, and V . Scarani, Phys. Rev. A 82, 012327 (2010)

work page 2010
[37]

Demianowicz and R

M. Demianowicz and R. Augusiak, Physical Review A 100 (2019), 10.1103/physreva.100.062318

work page doi:10.1103/physreva.100.062318 2019
[38]

Makuta, B

O. Makuta, B. Kuzaka, and R. Augusiak, Quantum 7, 915 (2023)

work page 2023
[39]

Y . S. Samoilenko, Mathematics and its Applications, V ol. 57 (Springer Science & Business Media Dordrecht, 1991)

work page 1991
[40]

Englbrecht, T

M. Englbrecht, T. Kraft, and B. Kraus, Quantum 6, 846 (2022)

work page 2022
[41]

H. F. Chau, Physical Review A 56, R1–R4 (1997)

work page 1997
[42]

L. Chen, M. Aulbach, and M. Hajdušek, Physical Review A 89 (2014), 10.1103/physreva.89.042305

work page doi:10.1103/physreva.89.042305 2014
[43]

Contreras-Tejada, C

P . Contreras-Tejada, C. Palazuelos, and J. I. de Vicente, Phys. Rev. Lett. 122, 120503 (2019)

work page 2019
[44]

K. V . Antipin,Journal of Physics A: Mathematical and Theoretical 54, 505303

work page
[45]

Optimizing strongly interacting fermionic hamiltonians,

M. B. Hastings and R. O’Donnell, “Optimizing strongly interacting fermionic hamiltonians,” (2023), arXiv:2110.10701 [quant-ph]

work page arXiv 2023
[46]

Kaniewski, I

J. Kaniewski, I. Šupi ´c, J. Tura, F. Baccari, A. Salavrakos, and R. Augusiak, Quantum 3, 198 (2019) . Appendix A: Proof of Theorem 1 Here, we formulate a proof of Theorem 1. To this end, we ﬁrst need to introduce and prove ﬁve necessar y lemmas that relate group A and generating graph g. Lemma 1 generalizes a result by Ref. [ 22] allowing us to transform...

work page 2019
[47]

Let us then focus on another pair of observables M3 andM4

Moreover, since Mi are unitary observables that satisfy M d i = /BD , it follows that M ′ i are also unitary which satisfy (M ′ i )d = /BD . Let us then focus on another pair of observables M3 andM4. It follows from the frustration matrix ( A1) that M3M4 =ω − 1M4M3, (A6) and so also M ′ 3M ′ 4 =ω − 1M ′ 4M ′

work page
[48]

Repeating this procedure m times produces a unitary U : H → (Cd)⊗ m ⊗ H ′ m, where H′ m is of unknown dimension, deﬁned as U :=U1(1d ⊗ U2)(1⊗ 2 d ⊗ U3)

(A7) Then, by the same argument as with M1,M 2 we have (1d ⊗ U2)M3(1d ⊗ U2)† = 1d ⊗ X2 ⊗ 1, (1d ⊗ U2)M4(1d ⊗ U2)† = 1d ⊗ Z2 ⊗ 1, (A8) whereU2 : H′ 1 → Cd ⊗ H ′ 2, and H′ 2 is a Hilbert space of an unknown dimension. Repeating this procedure m times produces a unitary U : H → (Cd)⊗ m ⊗ H ′ m, where H′ m is of unknown dimension, deﬁned as U :=U1(1d ⊗ U2)(1⊗...

work page
[49]

Let ei be ak-dimensional unit vector, i.e., it has a 1 entry on the i’th position and 0 elsewhere

we can rewrite it in terms of γ: k∑ i=1 k∑ j=1 IiJjγi,j = 0 (A12) for allJ ∈ Zk d and so in particular k∑ i=1 Iiγi,j = 0 (A13) for allj ∈ [k]. Let ei be ak-dimensional unit vector, i.e., it has a 1 entry on the i’th position and 0 elsewhere. Then we have γI = k∑ i,j=1 γj,iejeT i Iiei = k∑ j=1 k∑ i=1 Iiγj,iej = − k∑ j=1 k∑ i=1 Iiγi,jej = − k∑ j=1 0ej = 0, ...

work page

[1] [1]

Notice that {Ti}k i=1 is a generating set of A and the group operation in A is given byAI ⊙ AJ =AI+J , so thatI +J ∈ Zk d

such that • Ad I = 1 for anyI, • for any pair AI,A J ∈ A we have [AI,A J ]• =ω l1 for somel ∈ Zd. Notice that {Ti}k i=1 is a generating set of A and the group operation in A is given byAI ⊙ AJ =AI+J , so thatI +J ∈ Zk d. However, it is important to state that in this work, we are interested in commutation relations of elements A with respect to regular ma...

work page

[2] [2]

Theorem 2

is in fact bounded from above by ˜ω (G). Theorem 2. Let A = ⟨T1,...,T k⟩⊙ be a group as in Deﬁni- tion 1. F or each such A we have ∑ A∈A |⟨A⟩|2 ⩽d(null(γ)+k)/2 = ˜ω (G). (28) The detailed proof can be found in Appendix B. The main idea of the proof is to rewrite the sum of squares in Eq. ( 27) as ∑ A∈A |⟨A⟩|2 = tr [ ∑ A∈A (A ⊗ A†)ρ ⊗ ρ ] , (29) which then...

work page 2019

[3] [3]

M. T. Quintino, T. Vértesi, and N. Brunner, Phys. Rev. Lett. 113, 160402 (2014)

work page 2014

[4] [4]

Mayers and A

D. Mayers and A. Yao, in Proceedings 39th Annual Symposium on F oundations of Computer Science (Cat. No.98CB36280) (1998) pp. 503–509

work page 1998

[5] [5]

Zapatero, T

V . Zapatero, T. van Leent, R. Arnon-Friedman, W.-Z. Liu, Q. Zhang, H. Weinfurter, and M. Curty, npj Quantum Infor- mation 9, 10 (2023)

work page 2023

[6] [6]

D. P . Nadlinger, P . Drmota, B. C. Nichol, G. Araneda, D. Main, R. Srinivas, D. M. Lucas, C. J. Ballance, K. Ivanov, E. Y .-Z. Tan, P . Sekatski, R. L. Urbanke, R. Renner, N. Sangouard, and J.-D. Bancal, Nature 607, 682–686 (2022)

work page 2022

[7] [7]

Šupi ´c and J

I. Šupi ´c and J. Bowles, Quantum 4, 337 (2020)

work page 2020

[8] [8]

Baccari, R

F. Baccari, R. Augusiak, I. Šupi ´c, and A. Acín, Physical Review Letters 125 (2020), 10.1103/physrevlett.125.260507

work page doi:10.1103/physrevlett.125.260507 2020

[9] [9]

Acín and L

A. Acín and L. Masanes, Nature 540, 213 (2016). 8

work page 2016

[10] [10]

Pironio, A

S. Pironio, A. Acín, S. Massar, A. B. de la Giroday, D. N. Mat- sukevich, P . Maunz, S. Olmschenk, D. Hayes, L. Luo, T. A. Manning, and C. Monroe, Nature 464, 1021 (2010)

work page 2010

[11] [11]

Maccone and A

L. Maccone and A. K. Pati, Phys. Rev. Lett. 113, 260401 (2014)

work page 2014

[12] [12]

de Guise, L

H. de Guise, L. Maccone, B. C. Sanders, and N. Shukla, Phys. Rev. A 98, 042121 (2018)

work page 2018

[13] [13]

Y . Xiao, C. Guo, F. Meng, N. Jing, and M.-H. Y ung, Phys. Rev. A 100, 032118 (2019)

work page 2019

[14] [14]

Frustrated spin systems: History of the emergence of a mode rn physics,

H. T. Diep,“Frustrated spin systems: History of the emergence of a mode rn physics,” (2024), arXiv:2411.12826 [cond-mat.stat-mech]

work page arXiv 2024

[15] [15]

M. B. Hastings and R. O’Donnell, in Proceedings of the 54th Annual ACM SIGACT Symposium on Theor y of Computing , STOC 2022 (Association for Computing Machinery, New Y ork, NY , USA, 2022) p. 776–789

work page 2022

[16] [16]

Chapman and S

A. Chapman and S. T. Flammia, Quantum 4, 278 (2020)

work page 2020

[17] [17]

Chapman, S

A. Chapman, S. J. Elman, and R. L. Mann, “A uniﬁed graph-theoretic framework for free-fermion solv ability,” (2023), arXiv:2305.15625 [quant-ph]

work page arXiv 2023

[18] [18]

de Gois, K

C. de Gois, K. Hansenne, and O. Gühne, Phys. Rev. A 107, 062211 (2023)

work page 2023

[19] [19]

Z.-P . Xu, R. Schwonnek, and A. Winter, PRX Quantum 5 (2024), 10.1103/prxquantum.5.020318

work page doi:10.1103/prxquantum.5.020318 2024

[20] [20]

Full classiﬁcation of pauli lie algebras,

G. Aguilar, S. Cichy, J. Eisert, and L. Bit- tel, “Full classiﬁcation of pauli lie algebras,” (2024), arXiv:2408.00081 [quant-ph]

work page arXiv 2024

[21] [21]

Lovasz, IEEE Transactions on Information Theory 25, 1 (1979)

L. Lovasz, IEEE Transactions on Information Theory 25, 1 (1979)

work page 1979

[22] [22]

A graph-theoretic framework for free-parafermion solvab ility,

R. L. Mann, S. J. Elman, D. R. Wood, and A. Chapman, “A graph-theoretic framework for free-parafermion solvab ility,” (2024), arXiv:2408.09684 [quant-ph]

work page arXiv 2024

[23] [23]

Sarkar and T

R. Sarkar and T. J. Y oder,Quantum 8, 1307 (2024)

work page 2024

[24] [24]

Santos, C

R. Santos, C. Jebarathinam, and R. Augusiak, Phys. Rev. A 106, 012431 (2022)

work page 2022

[25] [25]

A. M. Steane, Phys. Rev. Lett. 77, 793 (1996)

work page 1996

[26] [26]

P . J. Nadkarni and S. S. Garani, Phys. Rev. A 103, 042420 (2021)

work page 2021

[27] [27]

Stabilizer Codes and Quantum Error Correction

D. Gottesman, “Stabilizer codes and quantum error correction,” (1997), arXiv:quant-ph/9705052 [quant-ph]

work page internal anchor Pith review Pith/arXiv arXiv 1997

[28] [28]

Tóth and O

G. Tóth and O. Gühne, Physical Review A 72 (2005), 10.1103/physreva.72.022340

work page doi:10.1103/physreva.72.022340 2005

[29] [29]

Tóth and O

G. Tóth and O. Gühne, Phys. Rev. A 72, 022340 (2005)

work page 2005

[30] [30]

Makuta and R

O. Makuta and R. Augusiak, New Journal of Physics 23, 043042 (2021)

work page 2021

[31] [31]

Sen(De) and U

A. Sen(De) and U. Sen, Phys. Rev. A 81, 012308 (2010)

work page 2010

[32] [32]

Seevinck and J

M. Seevinck and J. Ufﬁnk, Phys. Rev. A 65, 012107 (2001)

work page 2001

[33] [33]

SHIMONY ,Annals of the New Y ork Academy of Sciences 755, 675 (1995) https://nyaspubs.onlinelibrary.wiley.com/doi/pdf/10.1111/j.1749-6632.1995.tb3

A. SHIMONY ,Annals of the New Y ork Academy of Sciences 755, 675 (1995) https://nyaspubs.onlinelibrary.wiley.com/doi/pdf/10.1111/j.1749-6632.1995.tb3

work page doi:10.1111/j.1749-6632.1995.tb3 1995

[34] [34]

Barnum and N

H. Barnum and N. Linden, Journal of Physics A: Mathematical and General 34, 6787 (2001)

work page 2001

[35] [35]

Gour and N

G. Gour and N. R. Wallach, Phys. Rev. A 76, 042309 (2007)

work page 2007

[36] [36]

Branciard, H

C. Branciard, H. Zhu, L. Chen, and V . Scarani, Phys. Rev. A 82, 012327 (2010)

work page 2010

[37] [37]

Demianowicz and R

M. Demianowicz and R. Augusiak, Physical Review A 100 (2019), 10.1103/physreva.100.062318

work page doi:10.1103/physreva.100.062318 2019

[38] [38]

Makuta, B

O. Makuta, B. Kuzaka, and R. Augusiak, Quantum 7, 915 (2023)

work page 2023

[39] [39]

Y . S. Samoilenko, Mathematics and its Applications, V ol. 57 (Springer Science & Business Media Dordrecht, 1991)

work page 1991

[40] [40]

Englbrecht, T

M. Englbrecht, T. Kraft, and B. Kraus, Quantum 6, 846 (2022)

work page 2022

[41] [41]

H. F. Chau, Physical Review A 56, R1–R4 (1997)

work page 1997

[42] [42]

L. Chen, M. Aulbach, and M. Hajdušek, Physical Review A 89 (2014), 10.1103/physreva.89.042305

work page doi:10.1103/physreva.89.042305 2014

[43] [43]

Contreras-Tejada, C

P . Contreras-Tejada, C. Palazuelos, and J. I. de Vicente, Phys. Rev. Lett. 122, 120503 (2019)

work page 2019

[44] [44]

K. V . Antipin,Journal of Physics A: Mathematical and Theoretical 54, 505303

work page

[45] [45]

Optimizing strongly interacting fermionic hamiltonians,

M. B. Hastings and R. O’Donnell, “Optimizing strongly interacting fermionic hamiltonians,” (2023), arXiv:2110.10701 [quant-ph]

work page arXiv 2023

[46] [46]

Kaniewski, I

J. Kaniewski, I. Šupi ´c, J. Tura, F. Baccari, A. Salavrakos, and R. Augusiak, Quantum 3, 198 (2019) . Appendix A: Proof of Theorem 1 Here, we formulate a proof of Theorem 1. To this end, we ﬁrst need to introduce and prove ﬁve necessar y lemmas that relate group A and generating graph g. Lemma 1 generalizes a result by Ref. [ 22] allowing us to transform...

work page 2019

[47] [47]

Let us then focus on another pair of observables M3 andM4

Moreover, since Mi are unitary observables that satisfy M d i = /BD , it follows that M ′ i are also unitary which satisfy (M ′ i )d = /BD . Let us then focus on another pair of observables M3 andM4. It follows from the frustration matrix ( A1) that M3M4 =ω − 1M4M3, (A6) and so also M ′ 3M ′ 4 =ω − 1M ′ 4M ′

work page

[48] [48]

Repeating this procedure m times produces a unitary U : H → (Cd)⊗ m ⊗ H ′ m, where H′ m is of unknown dimension, deﬁned as U :=U1(1d ⊗ U2)(1⊗ 2 d ⊗ U3)

(A7) Then, by the same argument as with M1,M 2 we have (1d ⊗ U2)M3(1d ⊗ U2)† = 1d ⊗ X2 ⊗ 1, (1d ⊗ U2)M4(1d ⊗ U2)† = 1d ⊗ Z2 ⊗ 1, (A8) whereU2 : H′ 1 → Cd ⊗ H ′ 2, and H′ 2 is a Hilbert space of an unknown dimension. Repeating this procedure m times produces a unitary U : H → (Cd)⊗ m ⊗ H ′ m, where H′ m is of unknown dimension, deﬁned as U :=U1(1d ⊗ U2)(1⊗...

work page

[49] [49]

Let ei be ak-dimensional unit vector, i.e., it has a 1 entry on the i’th position and 0 elsewhere

we can rewrite it in terms of γ: k∑ i=1 k∑ j=1 IiJjγi,j = 0 (A12) for allJ ∈ Zk d and so in particular k∑ i=1 Iiγi,j = 0 (A13) for allj ∈ [k]. Let ei be ak-dimensional unit vector, i.e., it has a 1 entry on the i’th position and 0 elsewhere. Then we have γI = k∑ i,j=1 γj,iejeT i Iiei = k∑ j=1 k∑ i=1 Iiγj,iej = − k∑ j=1 k∑ i=1 Iiγi,jej = − k∑ j=1 0ej = 0, ...

work page