Pith. sign in

REVIEW 2 major objections 5 minor 1 cited by

Group Invariant Quantum Latin Squares

T0 review · 2 major / 5 minor · reviewed 2026-08-10 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A group-invariant quantum Latin square exists exactly when the two groups have the same multiset of irreducible representation degrees.

desk verdict Theorem 4.5's condition (2) drops a complex conjugate; the intended classification likely survives, but the printed statement and several proofs need correction. read the letter →

arxiv 2501.00196 v2 pith:TEATHQQG submitted 2024-12-31 math.QA math.COquant-ph

Arnbjörg Soffía Árnadóttir , David E. Roberson This is my paper · ORCID
classification math.QAmath.COquant-ph MSC 05C6020C0520C1505C2581P45
keywords quantumLatinsquaresgroupinvariantcorrelationsalgebrasunitaryisomorphismsgraphisomorphismCayleygraphsrepresentationdegreestransformationmatrices
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

The paper introduces a symmetry notion for quantum Latin squares with rows and columns indexed by finite groups G and G′: the inner product of the entries at (a,b) and (c,d) may depend only on $a^{{-1}}$c and $b^{{-1}}$d. Its aim is to classify these (G,G′)-invariant quantum Latin squares and to decide when they exist. The classification is complete up to global isometry: each such square is encoded by a transformation matrix, and these matrices are exactly the unitaries implementing trace- and conjugate-transpose-preserving isomorphisms between the group algebras of G and G′. It follows that a (G,G′)-invariant quantum Latin square exists exactly when the multisets of irreducible representation degrees of G and G′ coincide. The paper is explicit that this is a strengthening of group-invariant correlations, since only the squared moduli, not the full inner products, need be invariant for the motivating applications.

What carries the argument

The central object is the transformation matrix U, the |G|×|G′| matrix whose entries are the inner products U_{x,y} = ⟨ψ_{a,b}|ψ_{c,d}⟩ for $a^{{-1}}$c=x and $b^{{-1}}$d=y; it determines the quantum Latin square up to a global isometry. Theorem 4.5 characterizes these matrices by three conditions—unitarity, the symmetry U_{a,b}=U_{$a^{{-1}}$,$b^{{-1}}$}, and the convolution identity U_{ab,c}=\sum_{xy=c}U_{a,x}U_{b,y}. The argument turns on Corollary 6.5, which restates those conditions as Uλ′(b)U† = \sum_a U_{a,b}λ(a), so that U is exactly the unitary performing an isomorphism between the left regular representations. This is what connects the combinatorial Latin-square data to representation theory: the left regular representation spans the group algebra, whose isomorphism type is governed by the degrees of irreducible representations.

What would settle it

Compute a candidate (G,G′)-transformation matrix for groups whose irreducible representation degree multisets differ, such as S_3 and Z_6; Corollary 7.6 predicts no solution to the three conditions of Theorem 4.5, and finding a single such matrix would falsify the classification.

Watch

Extended reading notes

Core claim

The core discovery is that a (G,G′)-invariant quantum Latin square exists if and only if the group algebras Λ_G and Λ_{G′} are isomorphic as trace-preserving, conjugate-transpose-commuting algebras. In matrix form, a G×G′ matrix U is the transformation matrix of such a square precisely when the map Ψ_U(λ′(b)) = \sum_{a∈G} U_{a,b} λ(a) is a unitary isomorphism from Λ_{G′} to Λ_G, and every unitary isomorphism of the two group algebras arises from a unique such U. Since group algebras are semisimple, the Wedderburn–Artin theorem identifies them with direct sums of full matrix algebras, so this existence condition is equivalent to equality of the multisets of degrees of the irreducible representations. The same machinery shows that all (G,G′)-invariant quantum Latin squares can be constructed from block-diagonalizing unitaries, yielding finitely many squares in the abelian case and uncountably many in the non-abelian case.

Load-bearing premise

The load-bearing premise is that the group-invariance condition is imposed on the full inner products, not just their squared moduli, so the classification covers only a proper subclass of the group-invariant quantum correlations that motivate the study.

Editorial extensions

If this is right

  • If the classification is right, then the existence of a (G,G′)-invariant quantum Latin square is an equivalence relation on finite groups, and composing two such squares corresponds to multiplying their transformation matrices.
  • For abelian groups G and G′ of equal order, all transformation matrices are of the form C†P_πC′ for a permutation π of the character groups, where C and C′ are normalized character tables; in particular there are finitely many.
  • For non-abelian groups with isomorphic group algebras, there are uncountably many (G,G′)-invariant quantum Latin squares up to global isometry.
  • The correlation produced by a Z_2^4-invariant quantum Latin square can be strongly nonlocal, and the same construction lifts to Z_2^d for all d ≥ 4; this answers the question whether correlations from such squares are always classical.
  • Every (G,G′)-invariant quantum correlation wins the isomorphism game for exactly the pairs of Cayley (di)graphs whose connection sets are unions of the connected components of its support graph, so the support graph is a complete bookkeeping device for graph-isomorphism applications.

Reading between the lines

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

  • A practical consequence the paper does not spell out: the degree-multiset condition is a fast filter for the search for quantum-isomorphic Cayley graphs; candidate pairs can be restricted to groups with the same irreducible representation degrees, and among those with equal degrees the constructive recipe of Section 8 gives explicit strategies to test.
  • The paper's open projective-representation generalization would preserve group-invariant correlations while relaxing inner-product invariance to invariance up to phases; if it holds, the phenomenon of group-invariant correlations would be exactly quasi-regular projective representation theory.
  • For abelian groups, Theorem 9.4 reduces the question 'is this correlation non-classical?' to comparing two explicit polytopes, so the problem becomes a finite combinatorial search over permutations of the character group; the Z_2^4 counterexample shows the threshold lies at dimension four.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

2 major / 5 minor

Summary. The paper introduces (G,G')-invariant quantum Latin squares: quantum Latin squares indexed by finite groups G and G' whose inner products depend only on a^{-1}c and b^{-1}d. It defines the transformation matrix U of such a square, characterizes transformation matrices in Theorem 4.5, and then proves a structural classification: Theorem 7.5 identifies (G,G')-transformation matrices with trace- and conjugate-transpose-preserving isomorphisms of the group algebras Λ_G and Λ_{G'}, and Corollary 7.6 derives existence if and only if the multisets of degrees of irreducible representations of G and G' coincide. Sections 8 and 9 give explicit constructions using block-diagonalizing unitaries and, in the abelian case, character tables; Section 10 studies subgroups arising from sparse rows; Section 11 develops support graphs and decompositions; Section 12 presents computational examples, including a strongly nonlocal Z_2^4-invariant correlation and non-classical S_3-invariant correlations. The paper also connects these objects to quantum isomorphisms of Cayley (di)graphs via Theorems 3.17 and 3.18.

Significance. If the missing-conjugate issue identified below is fixed, this is a substantial contribution. The paper gives a complete classification of a natural class of highly symmetric quantum Latin squares up to global isometry, reduces existence to a very simple representation-theoretic criterion, and gives explicit constructions in the abelian and block-diagonalizable cases. The concrete computational examples are valuable and falsifiable: the Z_2^4 correlation matrix in Figure 2 can be checked directly, and the S_3 correlations are separated from the classical set by an explicit hyperplane. The paper also proves a useful structural result for support graphs and for the Cayley graphs for which a group-invariant correlation wins the isomorphism game. The authors are honest about the scope restriction: they classify inner-product-invariant quantum Latin squares, not the weaker and more general squared-modulus-invariant objects, and they state this limitation explicitly in Section 1.

major comments (2)
  1. [Theorem 4.5, Section 4] Condition (2) of Theorem 4.5 is printed without a complex conjugate and is false as stated. The condition satisfied by a (G,G')-invariant quantum Latin square is U_{a,b} = \overline{U_{a^{-1},b^{-1}}}, not U_{a,b} = U_{a^{-1},b^{-1}}. The forward proof uses the equality <ψ_{e,e}|ψ_{x,y}> = <ψ_{x,y}|ψ_{e,e}> without conjugating the inner product, and the paper's own Example 4.4 violates the printed condition: its entries α=(1-i)/2 and β=(1+i)/2 satisfy α = \bar{β}, not α = β. The converse construction also drops a conjugate in the expansion of <ψ_{a,b}|ψ_{c,d}>. With the conjugate inserted, the converse works using condition (3) with indices a^{-1} and c. This is load-bearing because Theorem 4.5 is the foundational characterization on which Sections 6 and 7 rely.
  2. [Lemma 6.1, Section 6] The proof of Lemma 6.1 contains the incorrect identity tr(ρ(g)ρ'(h)^†) = tr(ρ(g)^†ρ'(h)). These two traces are complex conjugates of one another, not equal in general. The correct conclusion of the calculation is U_{g^{-1},h^{-1}} = \overline{U_{g,h}}, which matches the corrected form of Theorem 4.5(2). Since Theorem 7.5 invokes Lemma 6.1, this part of the proof must be repaired together with the statement of Theorem 4.5.
minor comments (5)
  1. [Section 1, Remark 4.6, Section 13] The scope restriction to inner-product invariance rather than squared-modulus invariance is stated clearly, but it may deserve a more prominent caveat in the abstract: the resulting connection to group-invariant correlations covers only a proper subclass, and Open Problem 13.2 concerns non-classicality within this restricted class.
  2. [Lemmas 3.4 and 3.6] The proofs of Lemmas 3.4 and 3.6 are deferred to the authors' earlier paper [18]. The text says the generalization is straightforward, but since these lemmas are used throughout Section 3, including the proof of Theorem 3.17, a short proof or a precise statement of the two-group version would improve self-containedness.
  3. [Example 4.4] The displayed matrix for the (Z_4, Z_2 × Z_2) example would be easier to verify if the row/column ordering and the positions of α and β were explicitly tied to the group elements, since the corrected symmetry condition is immediately visible only when the indexing is made explicit.
  4. [Section 12.2] The non-classicality of the Z_2^4 example is found by a randomized search that is described in words but not accompanied by code, pseudocode, or a random seed. Because the explicit matrix is printed, the claim can be checked independently, but reproducibility would be improved by supplying the search procedure in more formal detail.
  5. [Throughout] There are a few minor typos (e.g., 'matarix' in Section 3.1). These do not affect the mathematics but should be cleaned up in revision.

Circularity Check

0 steps flagged · score 2.0 of 10

No significant circularity: the transformation-matrix classification and existence criterion are proven rather than assumed; the only self-citation is to prior correlation machinery and is not load-bearing for the main theorem.

full rationale

The paper's central claim (Theorem 7.5 and Corollary 7.6) is derived in a self-contained way. Theorem 4.5 directly characterizes transformation matrices by unitarity, the inner-product symmetry condition, and a convolution-type condition; its proof constructs the quantum Latin square from any matrix satisfying these conditions, so the characterization is not assumed by definition. Theorem 7.5 then shows that the linear map sending lambda'(b) to sum_a U_{a,b} lambda(a) is a unitary isomorphism exactly when U is a transformation matrix, using Corollary 6.5 and Lemma 6.1, which are themselves proved from the transformation-matrix axioms. Corollary 7.6 follows from the Wedderburn–Artin theorem and standard representation theory (Lemma 7.1, Lemma 7.3), not from any fitted or self-cited result. The paper does rely on the authors' earlier work [18] for definitions and correlation preliminaries (e.g., Lemma 3.4, Lemma 3.6, the correlation p_G), and explicitly says proofs are omitted because they are analogous; this is a genuine but minor self-citation. It is not load-bearing for the main transformation-matrix/isomorphism classification, which is proved here from first principles. The only notable issue found is a correctness erratum, not circularity: Theorem 4.5 condition (2) as printed omits a complex conjugate (the proof itself uses U_{a^{-1},b^{-1}} = overline{U_{a,b}}, and Example 4.4 violates the printed equality), and Lemma 6.1 has the same conjugate omission. This affects the formal statement but not the intended argument, and it does not change the circularity verdict.

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

The paper is pure mathematics: it introduces no fitted parameters and postulates no new physical entities. Its central claim rests on standard representation theory (Wedderburn-Artin, Schur's lemma, orthogonality relations) and on a disclosed domain assumption that the correct notion of symmetry is inner-product invariance rather than the weaker correlation invariance. No ad hoc constants or new objects with independent evidence requirements appear.

assumptions (4)
  • standard math Wedderburn-Artin theorem and the uniqueness of matrix algebra decomposition for semisimple Artinian algebras
    Invoked in Section 7 (Lemma 7.3) to equate group algebra isomorphism with matching multisets of irreducible representation degrees.
  • standard math Schur's lemma and character orthogonality relations for finite groups
    Used in Sections 6 and 7 to establish that irreducible representations span full matrix algebras and that the left regular representation is quasi-regular.
  • domain assumption The notion of (G,G')-invariance is imposed on inner products rather than on their squared moduli
    Section 1 states that characterizing quantum Latin squares with group-invariant correlations seems out of reach, so the paper strengthens invariance to the inner products themselves. All classification results (Theorems 4.5, 7.5, 8.4) apply only to this restricted class.
  • domain assumption For the application to quantum isomorphisms of Cayley graphs, it suffices to search among group-invariant correlations, and the quantum Latin squares constructed here form a subset of these
    Section 3.2 and Remark 4.6 note that every group-invariant quantum Latin square yields a group-invariant correlation, but not conversely. The relevance of the classification to the isomorphism game therefore only covers a proper subclass of group-invariant quantum correlations.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Group Invariant Quantum Latin Squares." pith.science (2026). https://pith.science/paper/TEATHQQG

@misc{pith2026250100196,
  author       = {Pith},
  title        = {Pith review of: Group Invariant Quantum Latin Squares},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TEATHQQG}},
  note         = {Machine review of arXiv:2501.00196}
}
abstract

A quantum Latin square is an $n \times n$ array of unit vectors where each row and column forms an orthonormal basis of a fixed complex vector space. We introduce the notion of $(G,G')$-invariant quantum Latin squares for finite groups $G$ and $G'$. These are quantum Latin squares with rows and columns indexed by $G$ and $G'$ respectively such that the inner product of the $a,b$-entry with the $c,d$-entry depends only on $a^{-1}c \in G$ and $b^{-1}d \in G'$. This definition is motivated by the notion of group invariant bijective correlations introduced in [Roberson \& Schmidt (2020)], and every group invariant quantum Latin square produces a group invariant bijective correlation, though the converse does not hold. In this work we investigate these group invariant quantum Latin squares and their corresponding correlations. Our main result is that, up to applying a global isometry to every vector in a $(G,G')$-invariant quantum Latin square, there is a natural bijection between these objects and trace and conjugate transpose preserving isomorphisms between the group algebras of $G$ and $G'$. This in particular proves that a $(G,G')$-invariant quantum Latin square exists if and only if the multisets of degrees of irreducible representations are equal for $G$ and $G'$. Another motivation for this line of work is that whenever Cayley graphs for groups $G$ and $G'$ are quantum isomorphic, then there is a $(G,G')$-invariant quantum correlation witnessing this, and thus it suffices to consider such correlations when searching for quantum isomorphic Cayley graphs. Given a group invariant quantum correlation, we show how to construct all pairs of graphs for which it gives a quantum isomorphism.

Figures

Figures reproduced from arXiv: 2501.00196 by the authors.

Figure 1
Figure 1. A (Z4,Z2 × Z2)-invariant quantum Latin square. Note that by definition of quantum Latin square, the transformation matrix of any group invariant quantum Latin square must of course be square. In fact, it turns out that the transformation matrix of a (G, G′ )-invariant quantum Latin square is unitary. Moreover, we prove the following: Theorem 4.5. Let G and G′ be finite groups. Then a matrix U ∈ C G×G′ is the transfo… view at source ↗
Figure 2
Figure 2. The correlation matrix of a strongly nonlocal quan [PITH_FULL_IMAGE:figures/full_fig_p050_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

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

  1. Large sets of mutually orthogonal quantum Latin squares

    math.CO 2026-07 unverdicted novelty 7.0 of 10

    A set of n−2 mutually orthogonal quantum Latin squares of order n must be classical, and for prime powers q the paper constructs d−1 of them, one non-classical, whenever d>1 divides q−1.

Reference graph

Works this paper leans on

21 extracted references · 17 canonical work pages · cited by 1 Pith paper

  1. [18]

    Roberson and Simon Schmidt

    David E. Roberson and Simon Schmidt. Quantum symmetry v s nonlocal symmetry

    work page

  2. [1]

    Minimum Quantum Resources for Strong Non- Locality

    Samson Abramsky, Rui Soares Barbosa, Giovanni Car` u, Na dish de Silva, Kohei Kishida, and Shane Mansfield. Minimum Quantum Resources for Strong Non- Locality. In 12th Conference on the Theory of Quantum Computation, Communica - tion and Cryptography (TQC 2017) , volume 73 of Leibniz International Proceedings in Informatics (LIPIcs) , pages 9:1–9:20, 2018....

    doi:10.4230/lipics.tqc.2017.9 2017

  3. [2]

    Roberson, R obert ˇS´ amal, Simone Severini, and Antonios Varvitsiotis

    Albert Atserias, Laura Manˇ cinska, David E. Roberson, R obert ˇS´ amal, Simone Severini, and Antonios Varvitsiotis. Quantum and non-sign alling graph iso- morphisms. Journal of Combinatorial Theory, Series B , 136:289 – 328, 2019. doi:10.1016/j.jctb.2018.11.002

    doi:10.1016/j.jctb.2018.11.002 2019

  4. [3]

    Design of comparative experiments , volume 25

    Rosemary A Bailey. Design of comparative experiments , volume 25. Cambridge University Press, 2008

    work page 2008

  5. [4]

    D. J. Benson. Representations and cohomology. I , volume 30 of Cambridge Studies in Advanced Mathematics . Cambridge University Press, Cambridge, second edition, 1998. 53

    work page 1998

  6. [5]

    On the orthogonal rank of ca yley graphs and im- possibility of quantum round elimination

    Jop Briet and Jeroen Zuiddam. On the orthogonal rank of ca yley graphs and im- possibility of quantum round elimination. Quantum Info. Comput. , 17(1–2):106–116, February 2017

    work page 2017

  7. [6]

    Magic squares: Latin, semiclassical, and quantum

    Gemma De las Cuevas, Tim Netzer, and Inga Valentiner-Bra nth. Magic squares: Latin, semiclassical, and quantum. Journal of Mathematical Physics , 64(2), 2023

    work page 2023

  8. [7]

    Algebraic graph theory , volume 207

    Chris Godsil and Gordon F Royle. Algebraic graph theory , volume 207. Springer Science & Business Media, 2001

    work page 2001

Show all 21 references
  1. [8]

    Representations and Characters of Groups

    Gordon James and Martin Liebeck. Representations and Characters of Groups . Cam- bridge University Press, New York, second edition, 2001

    2001

  2. [9]

    Roberson

    Martino Lupini, Laura Manˇ cinska, and David E. Roberson . Nonlocal games and quantum permutation groups. Journal of Functional Analysis , 279(5):108592, 2020. doi:10.1016/j.jfa.2020.108592

    2020

  3. [10]

    Maximally Entangled State in Pseudo-Telepathy Games, pages 200–207

    Laura Manˇ cinska. Maximally Entangled State in Pseudo-Telepathy Games, pages 200–207. Springer International Publishing, Cham, 2014. doi:10.1007/978-3-319-13350-8_15

    2014 doi

  4. [11]

    Roberson, and Antonios Varv isotis

    Laura Manˇ cinska, David E. Roberson, and Antonios Varv isotis. On deciding the existence of perfect entangled strategies for nonlocal gam es. Chicago Journal of The- oretical Computer Science , 2016(5), 2016. arXiv:1506.07429

    2016 arXiv

  5. [12]

    Roberson, and Antonios Varv itsiotis

    Laura Manˇ cinska, David E. Roberson, and Antonios Varv itsiotis. Graph isomor- phism: physical resources, optimization models, and algeb raic characterizations. Math. Program., 205(1):617–660, 2024. arXiv:http://arxiv.org/abs/2004.10893, doi:10.1007/S10107-023-01989-7

    2024 arXiv

  6. [13]

    Tracing the orbitals of the quantum permut ation group

    JP McCarthy. Tracing the orbitals of the quantum permut ation group. Archiv der Mathematik, 121(2):211–224, 2023

    2023

  7. [14]

    Constructing mutually unbiased bases from quantum latin squares

    Benjamin Musto. Constructing mutually unbiased bases from quantum latin squares. arXiv preprint arXiv:1605.08919 , 2016

    2016 arXiv

  8. [15]

    Quantum Latin squares and unitary er- ror bases

    Benjamin Musto and Jamie Vicary. Quantum Latin squares and unitary er- ror bases. Quantum Information & Computation , 16(15-16):1318–1332, 2016. doi:10.26421/QIC16.15-16

    2016 doi

  9. [16]

    Design of strong cryptographic schemes based on latin squar es

    Saibal K Pal, Shivam Kapoor, Alka Arora, Reshu Chaudhar y, and Jatin Khurana. Design of strong cryptographic schemes based on latin squar es. Journal of Discrete Mathematical Sciences and Cryptography , 13(3):233–256, 2010

    2010

  10. [17]

    Biunitary constructions in quan tum information

    D Reutter and J Vicary. Biunitary constructions in quan tum information. Higher Structures, 3(1), 2019

    2019

  11. [19]

    Linear representations of finite groups

    Jean-Pierre Serre. Linear representations of finite groups. , volume 42 of Graduate texts in mathematics . Springer, 1977. 54 A Proof of Lemma 8.3 Proof of Lemma 8.3. If U has the form given in the lemma statement, it is easy to see that U ˆAU † = ˆA. Conversely, suppose that U...

    1977

  12. [33]

    Clearly Ei ℓ,k has rank di and Ei ℓ,kEj s,t =δi,jδk,sEi ℓ,t

    such that Ai = 0 if i ⁄=j and Aj =eℓe† k. Clearly Ei ℓ,k has rank di and Ei ℓ,kEj s,t =δi,jδk,sEi ℓ,t. (35) Now choose ˆi ∈ [r] such that dˆi is minimum. Note that then Eˆi ℓ,k has rank dˆi which is the minimum nonzero rank of an element of ˆA. Since UE ˆi ℓ,kU † ∈ ˆA, we have...

  13. [34]

    If we let I i =∑ ℓ∈[di]Ei ℓ,ℓ, then I i is the identity in the subalgebra ˆAi

    to Ei ℓ,k, we have that ( UE i ℓ,kU † ) s,t =Usi ( Idi ⊗eℓe† k ) U † ti (36) 55 and by the above this is zero unless s =t =π(i). If we let I i =∑ ℓ∈[di]Ei ℓ,ℓ, then I i is the identity in the subalgebra ˆAi. From the above it is then easy to see that UI iU † = I π(i) and there...

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.