Pith. sign in

REVIEW 5 major objections 6 minor 50 references

Quantum strategies, error bounds, optimality, and duality gaps for multiplayer XOR, $\mathrm{XOR}^{*}$, compiled XOR, $\mathrm{XOR}^{*}$, and strong parallel repetiton of XOR, $\mathrm{XOR}^{*}$, and FFL games

T0 review · 5 major / 6 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read The paper claims that for multiplayer XOR games and their strong parallel repetitions, the quantum–classical duality gap is captured by a single semidefinite-program expression, and vanishes exactly when that expression is zero.

desk verdict The main theorem is a tautological restatement of SDP duality, and the N-player error bounds rest on a never-constructed operator; the paper should be desk-rejected rather than sent to referees. read the letter →

arxiv 2505.06322 v2 pith:3LORP6IV submitted 2025-05-09 quant-ph cs.ITmath.ITmath.PR

classification quant-phcs.ITmath.ITmath.PR MSC 81P0281Q02
keywords quantumgamesXORmultiplayernonlocalsemidefiniteprogrammingdualitygapserrorboundsstrongparallelrepetitionFFL
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

This paper tries to characterize when quantum strategies beat classical ones in games with more than two players: 3-XOR, 4-XOR, 5-XOR, N-XOR, and the strong parallel repetitions of XOR and FFL games. Its central claim is that, whenever the associated primal and dual semidefinite programs are well posed, the duality gap between the classical and quantum values is decided by a single condition, $(\sum_i y_i F_i - G)\cdot Z \geq 0$, with the gap vanishing exactly when this expression equals zero. The same collection of statements is extended to every game variant listed, and approximate optimality is certified by error bounds of the form $N!\, n^N\sqrt{\epsilon}$ up to constants. Such certificates matter because exact, checkable conditions on when entanglement helps would turn the question of quantum advantage in multiplayer games into a semidefinite-program computation.

What carries the argument

The load-bearing objects are the suitable linear operators $T_{\mathrm{3XOR}}$, $T_{\mathrm{N\,XOR}}$, and $T_{\mathrm{XOR}\wedge\cdots\wedge\mathrm{XOR}}$ — intertwining maps that move a player's observable from one tensor slot to another while reversing the order of the tensor product, for example $T_{\mathrm{3XOR}}:\mathbb{C}^{2\lceil n/3\rceil}\otimes\mathbb{C}^{2\lceil n/3\rceil}\otimes\mathbb{C}^{2\lceil n/3\rceil}\to \mathbb{C}^{d_A}\otimes \mathbb{C}^{d_B}\otimes \mathbb{C}^{d_C}$. Each is required to have unit Frobenius norm and to satisfy specified intertwining and anticommutation relations, and the error bounds and $\epsilon$-approximality inequalities are built on these relations. The second pillar is the semidefinite-program machinery: primal feasible solutions $Z$, symmetrized game tensors $G_{\mathrm{sym}}$, and dual variables $y_i$ with explicit combinatorial normalization, through which the duality gap is read off from $(\sum_i y_i F_i - G)\cdot Z$. The permutation collections $P_{\mathrm{3XOR}},\dots,P_{\mathrm{N\,XOR}}$ enumerate the ways player observables can be superposed, and the $\epsilon$-approximality inequalities bound the deviation of near-optimal strategies.

What would settle it

Explicitly construct the operator $T_{\mathrm{3XOR}}$ for a small instance, say three players with three questions each, and check the three intertwining identities written in Section 2.2 while directly computing its Frobenius norm; if no operator satisfying the identities has norm 1, every derived error bound and the vanishing-gap criterion collapse. Independently, one can solve the primal and dual SDPs for a small 3-XOR instance numerically and check whether the duality gap vanishes precisely when $(\sum_i y_i F_i - G)\cdot Z = 0$; the first failure would localize the defect to operator existence, the second to the SDP certificate itself.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is a unified semidefinite-program certificate for exact and approximate optimality across the family of multiplayer XOR-type games. Theorem 1 states that for the 3-XOR game the duality gap is captured by the condition $(\sum_{1\le i\le m} y_{\mathrm{3XOR},i} F_{\mathrm{3XOR},i} - G_{\mathrm{3XOR}})\cdot Z_{\mathrm{3XOR}} \ge 0$, that the gap vanishes if and only if this expression is zero, and that weak and strong duality are read off from the same expression; Theorems 2 through 6 assert the same collection of statements for the 4-XOR, 5-XOR, N-XOR, strong-parallel-repetition XOR, and strong-parallel-repetition FFL games. The dual variables are given in closed form with combinatorial weights such as $\omega_{\mathrm{NXOR}}/(N!\, n(n-1)\cdots(n-N+1))$, and the error bounds take the form $N!\, n^N\sqrt{\epsilon}$ up to constants. A contrast drawn in the paper: under strong parallel repetition the XOR optimal value multiplies as $\omega^n=(1/\sqrt{2})^n$, while the FFL value stays at $2/3$, so the FFL game has no duality gap and its gap-SDP primal feasible solution is identically zero.

Load-bearing premise

Every error bound and vanishing-gap statement in the paper rests on the existence of the intertwining operators $T_{\mathrm{3XOR}}$, $T_{\mathrm{N\,XOR}}$, and $T_{\mathrm{XOR}\wedge\cdots\wedge\mathrm{XOR}}$ with unit Frobenius norm and the stated anticommutation behavior — and Lemmas 9 and 10 establish the unit-norm property only by deferring to a two-player argument, without constructing the multiplayer operators.

Editorial extensions

If this is right

  • If the SDP certificate is valid, then for any well-posed N-player XOR-type game one can certify whether the quantum value provably beats the classical value by evaluating a single expression $(\sum_i y_i F_i - G)\cdot Z$; a zero value means strong duality holds and the SDP returns the exact quantum value.
  • The closed-form dual variables mean the certificate can be written down directly from the game tensor, with combinatorial weights $1/(N!\, n(n-1)\cdots(n-N+1))$, without solving an optimization problem to find the dual.
  • The error bound $N!\,n^N\sqrt{\epsilon}$ up to constants quantifies near-optimality: a strategy whose observables are $\epsilon$-close in Frobenius norm wins with probability within a factor $(1-\epsilon)$ of the optimal value, uniformly over the number of players.
  • For strong parallel repetition of the FFL game the classical and quantum values coincide at $2/3$, so the duality-gap SDP has an identically vanishing primal feasible solution; the game admits no quantum advantage under repetition, in contrast to XOR games whose optimal value multiplies as $(1/\sqrt{2})^n$.

Reading between the lines

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

  • If the unit-norm intertwining operators exist as assumed, the same certificate strategy should transfer to other permutation-symmetric multiplayer games whose tensors admit the same superposition structure, for instance multiplayer CHSH-type or linear games; this is a testable extension the paper does not itself pursue.
  • The $N!$ factor in the error bound suggests that writing down certificates for large $N$ is combinatorially expensive in general, and exploiting the symmetrization that produces $G_{\mathrm{sym}}$ may be the only practical route, making the certificate's true cost depend on the game's symmetry rather than its player count.
  • A direct numerical test of the paper's main theorem is available: for a small 3-XOR instance, solve the primal and dual SDPs independently and check that the gap vanishes precisely when the expression $(\sum_i y_i F_i - G)\cdot Z$ equals zero; failure on any instance would localize the defect without settling the operator-existence question.
  • The XOR-versus-FFL contrast under strong parallel repetition suggests a broader pattern: games without a duality gap stay gap-free under repetition because their value is pinned, while games with a gap amplify the gap multiplicatively; the paper's framework offers a template for asking which games behave which way.
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

5 major / 6 minor

Summary. The paper claims to characterize exact and approximate optimality for multiplayer (3-, 4-, 5-, and N-player) XOR games and for strong parallel repetitions of XOR and FFL games. The announced results include semidefinite programs for primal feasible solutions and duality gaps, error bounds of the form N! n^N sqrt(epsilon) (up to constants) for N-player XOR games and their strong parallel repetitions, positive-semidefiniteness of the associated dual objectives, and unit-Frobenius-norm "suitable linear operators" T_{3XOR}, T_{N XOR}, T_{XOR and ... and XOR}, and T_{FFL and FFL}. The central theorem (Theorem 1, Section 1.5) states a vanishing-duality-gap condition, a weak-duality characterization, and SDP duality statements; Theorems 2-6 assert the same collection of items for 4-XOR, 5-XOR, N-XOR, strong parallel repetition of XOR, and strong parallel repetition of FFL. The remaining results (Theorems 1*-8* and Lemmas T-1 through T-8) either reduce to the same template or are stated with one-sentence proofs.

Significance. The target of the paper, extending Ostrev's two-player error-bound framework [37] to multiplayer XOR-type games and to strong parallel repetitions, is a reasonable research direction, and the observation that strong parallel repetition affects XOR and FFL games differently (omega_{FFL and FFL} = 2/3 while omega_{XOR and XOR} = (omega_{XOR})^2, Section 1.4) is worth drawing attention to. The combinatorial enumeration of permutations of player observables in Tables 1-10 shows genuine effort. However, the paper ships no machine-checked proofs, no reproducible code, and no parameter-free derivation: the main theorem is a restatement of definitions, the dual variables are written down by hand, the positive-semidefiniteness proofs fix free constants after the fact, and every quantitative claim rests on "suitable linear operators" that are never constructed for N >= 3. As it stands, the paper does not establish any new verifiable quantitative claim beyond the two-player results already in [37] and [44].

major comments (5)
  1. [Section 1.5, Theorem 1] The third bullet of Theorem 1 states that weak duality, v_Primal,3XOR <= v_Dual,3XOR, holds "iff" (sum_i y_{3XOR,i} F_{3XOR,i} - G_{3XOR}) . Z_{3XOR} != 0. Weak duality for a feasible primal-dual SDP pair always holds, and it is not characterized by the duality-gap expression being nonzero. The preceding bullet ("Vanishing duality gap ... iff expression == 0") is a tautology, because the duality gap was defined one bullet earlier as that same expression being nonnegative. Since Theorems 2-6 repeat verbatim "the same collection of items" as Theorem 1, the six main theorems contain no content beyond their own definitions.
  2. [Section 1.5, Lemma T-1, Lemmas T-2 and T-3] The proof of Lemma T-1 concludes that the operator sum_i y_{N XOR,i} E_{N XOR,ii} - G_{N XOR,Sym} is positive semidefinite "from the observation that taking the constant C_{N XOR}, in the normalization ... to equal N! implies the desired result." The coefficients y_i were fixed in Section 1.5 to be omega_{N XOR}/(N! n), omega_{N XOR}/(N! n(n-1)), and so on, so the dual feasibility constraint is enforced by choosing the free constant after the fact rather than by verifying that the stated y's satisfy sum_i y_i E_ii >= G_Sym. The same pattern is repeated in Lemma T-2 (C = 3!) and Lemma T-3 (C = 4!). No argument shows that the chosen constants are legitimate or that dual optimality follows; the proof of the load-bearing PSD claim is circular.
  3. [Sections 2.4-2.5, Lemmas 9 and 10] Lemma 9 asserts that the Frobenius norm of the "suitable linear operators" for the 3-XOR, 4-XOR, 5-XOR, and N-XOR games equals 1, and Lemma 10 makes the same assertion for strong parallel repetition and for the two-player FFL game; both proofs consist solely of "Directly apply the argument from 6.2 in [37]." The operators T_{3XOR}, T_{N XOR}, T_{XOR and ... and XOR}, and T_{FFL and FFL} are never defined: domains and codomains are stipulated, but the maps themselves, their action on the optimal states, their intertwining relations with the player observables, and their anticommutation behavior are not given. The argument in [37, Section 6.2] constructs T only for two players, and the manuscript gives no reason that it extends to N >= 3, where the tensor-product structure is qualitatively different. Because Theorem 1* and Lemmas 1-3XOR, 1-N-XOR, Gen-FFL-Bound, and FR and ... and FR all depend on ||T||_F = 1 and on these intertwining relations, the error bounds have no foundation as written.
  4. [Section 2.3.1, Lemma 1-N-XOR proof; Sections 2.6 and 2.3.2, Lemmas 5*, 5**, Gen-FFL-Bound] The error bounds contain the target optimal value inside the bound. The proof of Lemma 1-N-XOR bounds ||...||_F < (n_1 + (n_1 + 2) omega^{-1}_{N XOR}) n^N sqrt(epsilon), and Lemma 5* and Lemma 5** contain omega^3_{N XOR} and omega_{XOR and ... and XOR}, respectively, inside the stated upper bounds. Such bounds cannot certify approximate optimality without prior knowledge of the very quantity the framework is meant to characterize, and they are not of the claimed "up to constants" form unless omega is already known. In addition, Lemma 2 ("N N-XOR identities") proves the identity by tacitly assuming a product structure T = T_XOR tensor I or T tensor T tensor ... tensor T inside the tensor product, which is precisely the structure whose existence is at issue for N >= 3.
  5. [Section 2.3.3, Theorems 4*-8* and Lemmas T-4 through T-8] Each of the five strong-parallel-repetition theorems and the five associated positive-semidefiniteness lemmas has a proof that reads only "Apply the same computations as provided in Theorem 1*, Theorem 2*, and Theorem 3*" or "Apply the same computations as provided in Lemma T-1, Lemma T-2, and Lemma T-3." No computation is shown for the parallel-repetition setting, and since the underlying operator T_{XOR and ... and XOR} is unconstructed (see the major comment on Lemmas 9 and 10), these one-sentence proofs do not establish the claims.
minor comments (6)
  1. [Title and throughout] The manuscript contains numerous typos and broken notation, including "repetiton" in the title, "referree," "Prinmal," "ineqaualities," "approximality," "soem," and "odrinary," together with inconsistent renderings of omega_{N XOR}, omega(N XOR), and omega_N XOR.
  2. [Section 1.5, y_{N XOR} listing] The indexing of the dual variables is garbled: the last line of the y_{N XOR} block repeats "y_{N XOR,1} == ... == y_{N XOR,n}" instead of continuing to indices n^{N-1}+1 through n^N, and the 5-XOR block has a line "y_{5XOR,n3+1} == ... == y_{5XOR,n4}" that is missing the "==" sign and inconsistent with the surrounding pattern.
  3. [Section 2.5, Lemma 10] The statement of Lemma 10 appears to be mis-copied from Lemma 9: although the lemma is titled for strong parallel repetition of the multiplayer XOR game and the two-player FFL game, its sentence lists exactly the same games as Lemma 9 ("3-XOR, 4-XOR, 5-XOR, and N-XOR").
  4. [Sections 1.4 and 2.6] The values omega_{XOR and XOR} = (omega_{XOR})^2 = 1/2 and omega_{FFL and FFL} = 2/3 are asserted repeatedly without proof or citation; since the FFL and XOR optimal values under strong parallel repetition are used inside the subsequent error-bound inequalities, a derivation or reference would be needed.
  5. [Section 2.3.3, proof of Lemma 5B**] The proof of Lemma 5B** contains an unexplained token "~www~" and a chain of steps labeled with "approx" in which quantities such as ||pm B_{kl} + B_{lk}||/sqrt(2) are replaced by 1; these are heuristic estimates, not rigorous inequalities, and the displayed manipulations do not yield the claimed bound 20N sqrt(N epsilon^ and).
  6. [References] The paper contains no bibliography: all citations are numeric placeholders ([37], [44], etc.), which makes it impossible to verify which statements are taken from the prior literature and which are claimed as new.

Circularity Check

4 steps flagged · score 8.0 of 10

N-player XOR bounds rest on hand-chosen dual variables, cited-but-unconstructed unit-norm operators T_{N XOR}, and 'apply the same computations' proofs; Theorem 1 restates SDP duality by definition.

  1. renaming known result [Section 1.5, Theorem 1 (extended by Theorems 2-6)]
    "The duality gap, which captures the difference between classical and quantum values of a game, is captured through the condition, (Σ_{1≤i≤m} y_{3XOR,i}F_{3XOR,i}−G_{3XOR})·Z_{3XOR}≥ 0 ... The duality gap formulated in the previous item above vanishes, v_{Primal,3XOR}≡v_{Dual,3XOR}, iff, (Σ y_{3XOR,i}F_{3XOR,i}−G_{3XOR})·Z_{3XOR}≡ 0."

    This 'characterization' is the textbook definition of SDP weak duality and complementary slackness, restated as a theorem. For any feasible primal Z and dual y, v_D = Σ_i y_i (F_i·Z), so (Σ y_i F_i − G)·Z = v_D − v_P. Weak duality says this quantity is ≥ 0, and it is zero exactly when v_P = v_D, i.e., exactly when the duality gap vanishes. The theorem's first and second bullets are therefore tautologically true by the definitions of the primal, dual, and duality gap; no game-specific derivation is supplied. The same definitional restatement is then copied verbatim into Theorems 2-6 for 4-XOR, 5-XOR, N-XOR, and parallel repetitions.

  2. ansatz smuggled in via citation [Sections 2.4-2.5, Lemma 9 and Lemma 10]
    "Lemma 9 (the Frobenius norm of suitable linear operators for the 3-XOR, 4-XOR, 5-XOR, and N-XOR games equals 1). ... Proof of Lemma 9. Directly apply the argument from 6.2 in [37], from which we conclude the argument."

    The operators T_{3XOR}, T_{4XOR}, T_{5XOR}, T_{N XOR} are never constructed: no map, domain-to-codomain action, intertwining relations, or anticommutation relations are exhibited beyond a stated domain of the form ⊗^N C^{2^{⌈n/2⌉}} → ⊗ C^{d_i}. Section 6.2 of [37] builds a two-player T only; Lemma 9 claims the N-player unit-norm property by 'directly apply the argument' with no mechanism for additional players. Every quantitative error bound in the paper (Lemma 1-N-XOR, Lemma 1−3XOR, Lemma 1-XOR strong parallel repetition, Lemma FFL strong parallel repetition) has the shape ‖(player observable)T − T(player observable)^~‖ < c n^N√ε and so presupposes the existence and unit-Frobenius norm of this T.

2 more flagged steps
  1. fitted input called prediction [Section 1.5 dual variables and Lemma T-1 proof (with Lemma T-2, T-3)]
    "y_{N XOR,1}≡···≡ y_{N XOR,n}≡ω_{N XOR}(1/(N! n)), ... the associated operator is positive semidefinite from the observation that taking the constant C_{N XOR}, in the normalization, 1/(C_{N XOR}ω_{N XOR}) n(∏_{1≤j≤N−1}(n−j)), to equal N! implies the desired result."

    The N! factor that appears in the announced error bound N! n ∏(n−j) is placed into the hand-written dual variables y_i (denominator N!) and then the PSD proof simply sets the free constant C_{N XOR} to N! to 'imply the desired result.' The choice of the dual solution and the free constant is therefore tailored to make the symmetrized operator PSD and the bound hold by construction, rather than the bound emerging from an independent argument. This is a fitted parameter renamed as a prediction: the paper's central combinatorial factor is an input selected to force the conclusion.

  2. self citation load bearing [Section 2.3.3, Theorem 4* through Theorem 8* (and Lemma T-4 through Lemma T-8)]
    "Theorem 4∗ (3-XOR strong parallel repetition error bounds, 2.2.1, Theorem 4, [37], Theorem 2, [44], Theorems 1-6 in 1.5). ... Proof of Theorem 4∗. Apply the same computations as provided in Theorem 1∗, Theorem 2∗, and Theorem 3∗, from which we conclude the argument."

    The central strong-parallel-repetition results are asserted to follow by 'apply the same computations' from the paper's own earlier theorems, with no computation, new operator, or independent estimate supplied. Those earlier theorems themselves rely on the unconstructed unit-norm operators of Lemma 9-10 and on the hand-chosen dual variables of Section 1.5. The proof chain for strong parallel repetition is therefore a self-citation loop: Theorem 4* points to Theorem 1*-3*, which point back to the paper's own asserted decompositions and to [44], without any verifiable derivation of the claimed N! n^N√ε-style bounds.

full rationale

The paper's headline claim, Theorem 1, is a direct transcription of SDP weak-duality/complementary-slackness: (Σ y_i F_i − G)·Z = v_D − v_P, so 'gap vanishes iff this expression is zero' is true by definition, not by game-theoretic content; Theorems 2-6 merely repeat that definitional statement for other games. The genuinely quantitative content—N-player Frobenius-norm bounds of order N! n^N√ε—is not self-contained: Lemma 9-10 assert unit Frobenius norm for multiplayer operators T_{3XOR}, T_{N XOR}, T_{XOR∧...∧XOR} with proofs that only say 'Directly apply the argument from 6.2 in [37],' while the operators themselves are never constructed for N ≥ 3. The dual variables y_i are written by hand with 1/(N!) prefactors, and the PSD proofs set the free constant C_{N XOR} to N! to obtain the desired result, so the N! factor is fitted, not derived. Strong-parallel-repetition theorems are proved by 'Apply the same computations' from the paper's own unproven earlier theorems, forming a self-citation chain rather than independent support. No external benchmark, machine-checked proof, or parameter-free falsifiable test is given. These features are not mere self-citation: they mean the central quantitative claims reduce by construction to the chosen ansatz, the chosen constants, and definitional SDP identities. Score 8 is warranted because the core error-bound claims are forced by the paper's own inputs and cited-but-unconstructed operators, though some peripheral content (permutation enumerations, Bell-state lists) is non-circular.

Assumptions & free parameters 4 free parameters · 5 assumptions · 2 invented entities

The central claim rests on prior-work operator existence, an assumed SDP setup, and hand-chosen constants; these are not independently established.

free parameters (4)
  • Arbitrary constants c_i, C_i, C'_N, C^and_2, C^and_N = chosen after the fact, e.g., C = C' = 5(N n^N)^2 sqrt(epsilon) in Lemma 1-N XOR
    In Lemma 1-N XOR and the strong parallel repetition analogs, constants are introduced and then set to make the stated inequalities close; no independent value is derived.
  • Dual variables y_{kXOR,i} = claimed formulas such as omega_{N XOR}/(N! n(n-1)...)
    These are announced in Section 1.5 without derivation and then used to enforce positive semidefiniteness by choosing normalization constants.
  • Optimality slack epsilon and starred variants = not specified; sufficiently small
    Every bound is an up-to-constants sqrt(epsilon) statement for sufficiently small epsilon; the text never fixes a concrete epsilon relation, so the bounds are not quantitatively checkable.
  • Game optimal value omega_{N XOR} = not provided for general N
    The error bounds are written in terms of omega^{-1}_{N XOR}; for general N the paper does not supply this value, making the bounds conditional.
assumptions (5)
  • domain assumption Primal and dual SDPs for each game are well posed and admit primal feasible solutions.
    Theorem 1 begins with this supposition and Theorems 2 through 6 inherit it; no proof of well-posedness or feasibility is given.
  • ad hoc to paper Suitable linear operators T_{3XOR}, T_{N XOR}, T_{XOR and ... and XOR} exist, have unit Frobenius norm, and satisfy intertwining and anticommutation relations.
    Lemmas 9 and 10 assert unit norm by directly applying the argument from 6.2 in [37]; the multiplayer intertwining identities are assumed throughout Section 2.
  • ad hoc to paper The N-th player tensor observable decomposes recursively as (P_{N-1}(i_1,...,i_{N-1}) + P_{N-1}(i'_1,...,i'_{N-1}))/sqrt(2).
    Stated in Section 2.6 without proof; this decomposition is used to generate the family of error-bound inequalities.
  • domain assumption The optimal value of XOR games multiplies under strong parallel repetition, and the FFL value stays 2/3.
    Used in Sections 1.4 and 2.6 to compute products and powers of omega; the paper treats these identities as known without derivation.
  • standard math Standard SDP weak duality and strong duality apply to the formulated programs.
    Invoked in Theorem 1; mathematically standard for feasible SDPs, but the paper's own weak-duality characterization is not standard and is stated incorrectly.
invented entities (2)
  • Suitable linear operators T_{3XOR}, T_{N XOR}, T_{XOR and ... and XOR}
    purpose: To generate error bounds by permuting and intertwining tensor observables of N players.
    No explicit construction is given; the unit-Frobenius-norm property is asserted by reference to the two-player proof, and no falsifiable prediction or external benchmark is offered.
  • N-player tensor observable decomposition family P_N(i_1,...,i_N)
    purpose: To express every player's response operator as a normalized superposition of lower-player observables.
    It is introduced and used to derive inequalities, but its existence for arbitrary N is not proven.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Quantum strategies, error bounds, optimality, and duality gaps for multiplayer XOR, $\mathrm{XOR}^{*}$, compiled XOR, $\mathrm{XOR}^{*}$, and strong parallel repetiton of XOR, $\mathrm{XOR}^{*}$, and FFL games." pith.science (2026). https://pith.science/paper/3LORP6IV

@misc{pith2026250506322,
  author       = {Pith},
  title        = {Pith review of: Quantum strategies, error bounds, optimality, and duality gaps for multiplayer XOR, $\mathrmXOR^*$, compiled XOR, $\mathrmXOR^*$, and strong parallel repetiton of XOR, $\mathrmXOR^*$, and FFL games},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/3LORP6IV}},
  note         = {Machine review of arXiv:2505.06322}
}
abstract

We characterize exact, and approximate, optimality of games that players can interact with using quantum strategies. In comparison to a previous work of the author, arXiv: 2311.12887, which applied a 2016 framework due to Ostrev for constructing error bounds beyond CHSH and XOR games, in addition to the existence of well-posed semidefinite programs for determining primal feasible solutions, along with quantum-classical duality gaps, it continues to remain of interest to further develop the construction of error bounds, and related objects, to game-theoretic settings with several participants. In such settings, one encounters a rich information theoretic landscape, not only from the fact that there exists a significantly larger combinatorial space of possible strategies for each player, but also several opportunities for pronounced quantum advantage. We conclude this effort by describing other variants of other possible strategies, as proposed sources for quantum advantage, in $\mathrm{XOR}^{*}$, compiled $\mathrm{XOR}^{*}$, and strong parallel repetition variants of $\mathrm{XOR}^{*}$ games.

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

50 extracted references · 2 canonical work pages

  1. [37]

    ACM Transactions on Computation Theory 4(15) https://doi.org/10.1145/2799560 205

    Oded, R., Vidick, T.: Quantum xor games. ACM Transactions on Computation Theory 4(15) https://doi.org/10.1145/2799560 205

  2. [44]

    arXiv: 2311.12887 (submitted) (2023)

    Rigas, P.: Optimal, and approximately optimal, quantum strategies for XOR ∗ and FFL games. arXiv: 2311.12887 (submitted) (2023)

  3. [1]

    classical two way communication in xor games

    Amr, A., Villanueva, I.: Quantum one way vs. classical two way communication in xor games. Quantum Information Processing 20(79) (2021) 202

  4. [2]

    STACS 12, 1–12 (2019) https://doi.org/10.4230/LIPIcs.STACS.2019.12

    Bannik, T., al.: Bounding quantum-classical separations for classes of nonlocal games. STACS 12, 1–12 (2019) https://doi.org/10.4230/LIPIcs.STACS.2019.12

  5. [3]

    Briet, J., Buhrman, H., Toner, B.: A generalized grothendieck inequality and entanglement in xor games. Comm. Math. Phys. 305, 827–843 (2011) https:// doi.org/10.1007/s00220-011-1280-3

  6. [4]

    Theoretical Computer Science 358, 3–14 (2006) https://doi.org/10.1016/j.tcs.2005.08.035

    Broadbent, A., Methot, A.A.: On the power of non-local boxes. Theoretical Computer Science 358, 3–14 (2006) https://doi.org/10.1016/j.tcs.2005.08.035

  7. [5]

    Brassard, G., Broadbent, A., Tapp, A.: Quantum pseudo-telepathy. Found. Phys. 35, 1877–1907 (2005) https://doi.org/https://philpapers.org/rec/BRAQP

  8. [6]

    Phys Rev Applied16(044057) (2021) https://doi.org/10.1103/PhysRevApplied.16.044057

    Benedetti, M., Coyle, B., Fiorentini, M., Lubasch, M., Rosenkranz, M.: Varia- tional Inference with a Quantum Computer. Phys Rev Applied16(044057) (2021) https://doi.org/10.1103/PhysRevApplied.16.044057

Show all 50 references
  1. [7]

    Phys- ical Review Letters 127(120502) (2021) https://doi.org/10.1103/PhysRevLett

    Bittel, L., Kliesch, M.: Training variational quantum algorithms is np-hard. Phys- ical Review Letters 127(120502) (2021) https://doi.org/10.1103/PhysRevLett. 127.120502

  2. [8]

    Catani, L., Faleiro, R., Emeriau, P.E., Mansfield, S., Pappa, A.: Connecting xor and xor* games. Phys. Rev. A 109(012427) (2024) https://doi.org/10.1103/ PhysRevA.109.012427

  3. [9]

    Physical Review Research 4(043119) (2022) https://doi

    Chen, H., Vives, M., Metcalf, M.: Parametric amplification of an optomechanical quantum interconnect. Physical Review Research 4(043119) (2022) https://doi. org/10.1103/PhysRevResearch.4.043119

  4. [10]

    New Journal of Physics 18(073011) (2016) https://doi.org/10

    Cong, I., Duan, L.: Quantum discriminant analysis for dimensionality reduction and classification. New Journal of Physics 18(073011) (2016) https://doi.org/10. 1088/1367-2630/18/7/073011

  5. [11]

    19th IEEE Annual Conference on Computational Complexity Proceedings, 236–249 (2004) https://doi.org/10.1109/CCC.2004.1313847

    Cleve, R., Hoyer, P., Toner, B., Watrous, J.: Consequences and limits of non- local strategies. 19th IEEE Annual Conference on Computational Complexity Proceedings, 236–249 (2004) https://doi.org/10.1109/CCC.2004.1313847

  6. [12]

    IEEE 65th Annual Symposium on Foundations of Computer Science (FOCS), 920–929 (2024) https://doi.org/10.1109/FOCS61266.2024.00061

    Culf, E., Mousavi, H., Spirig, T.: Approximation algorithms for noncommuta- tive csps. IEEE 65th Annual Symposium on Foundations of Computer Science (FOCS), 920–929 (2024) https://doi.org/10.1109/FOCS61266.2024.00061

  7. [13]

    arXiv: 2402.17301 (2024)

    Cui, D., Malavolta, G., Mehta, A., Natarajan, A., Paddock, C., Schmidt, S., Walter, M., Zhang, T.: A computational tsireslson’s theorem for the value of compiled xor games. arXiv: 2402.17301 (2024)

  8. [14]

    23rd Annual IEEE Conference on 203 Computational Complexity 8 (2018)

    Doherty, A.C., Liang, Y.C., Toner, B., Wehner, S.: The quantum moment problem and bounds on entangled multi-prover games. 23rd Annual IEEE Conference on 203 Computational Complexity 8 (2018)

  9. [15]

    Srinivas, Cabello, A., al.: Experimental quantum advantage in the odd- cycle game

    Drmota, P., Main, D., Ainley, E.M., Agrawal, A., Araneda, G., Nadlinger, R. Srinivas, Cabello, A., al.: Experimental quantum advantage in the odd- cycle game. Phys. Rev. Lett. 134(070201) (2025) https://doi.org/10.1103/ PhysRevLett.134.070201

  10. [16]

    IEEE Transactions on Microwave Theory and Techniques 70(5), 2517–2525 (2022) https://doi.org/10.1109/TMTT.2022

    Ewe, W.-B., Koh, D.E., Goh, S.T., Chu, H.-S., Png, C.E.: Variational quantum- based simulation of waveguide modes. IEEE Transactions on Microwave Theory and Techniques 70(5), 2517–2525 (2022) https://doi.org/10.1109/TMTT.2022. 3151510

  11. [17]

    PRX Quantum 3(020307) (2022) https://doi.org/ 10.1103/PRXQuantum.3.02030

    Pierre-Emmanuel Emeriau, P.-E., Howard, M., Mansfield, S.: Quantum advan- tage in information retrieval. PRX Quantum 3(020307) (2022) https://doi.org/ 10.1103/PRXQuantum.3.02030

  12. [18]

    Quantum Inf Process 19(229) (2020) https://doi.org/10.1007/s11128-020-02717-

    Faleiro, R.: Quantum strategies for simple 2-player xor games. Quantum Inf Process 19(229) (2020) https://doi.org/10.1007/s11128-020-02717-

  13. [19]

    Advance in Neural Information Processing Systems 32 (2019)

    Garg, D., Ikbal, S., Srivastava, S.K., Vishwakarma, H., Karanam, H., Subrama- niam, L.V.: Quantum embedding of knowledge for reasoning. Advance in Neural Information Processing Systems 32 (2019)

  14. [20]

    Jour- nal of Physics A: Mathematical and Theoretical 52(43) (2019) https://doi.org/ 10.1088/1751-8121/ab3fe0

    Genoni, M.G., Tufarelli, T.: Non-orthogonal bases for quantum metrology. Jour- nal of Physics A: Mathematical and Theoretical 52(43) (2019) https://doi.org/ 10.1088/1751-8121/ab3fe0

  15. [21]

    Phys.Rev.A 108(032409) (2023) https://doi.org/10.1103/ PhysRevA.108.032409

    Gidi, J.A., Candia, B., Munoz-Moller, A.D., Rojas, A., Pereira, L., Munoz, M., Zambrano, L., Delgado, A.: Stochastic optimization algorithms for quan- tum applications. Phys.Rev.A 108(032409) (2023) https://doi.org/10.1103/ PhysRevA.108.032409

  16. [22]

    AIAA 58(8) (2020)

    Givi, P., Daley, A.J., Mavriplis, D., Malik, M.: Quantum speedup for aeroscience and engineering. AIAA 58(8) (2020)

  17. [23]

    Helton, J.W., Mousavi, H., Nezhadi, S.S., al.: Synchronous values of games. Ann. Henri Poincar´ e 25, 4357–4397 (2024) https://doi.org/10.1007/ s00023-024-01426-1

  18. [24]

    IEEE Transactions on Information Theory 70(3), 1876– 1896 (2024) https://doi.org/10.1109/TIT.2023.3324527

    Hadiashar, S.B., Nayak, A., Sinha, P.: Optimal lower bounds for quantum learning via information theory. IEEE Transactions on Information Theory 70(3), 1876– 1896 (2024) https://doi.org/10.1109/TIT.2023.3324527

  19. [25]

    Quantum Machine Intelligence 4(3) (2022) https://doi.org/ 10.1007/s42484-021-00061-x 204

    Hur, T., Kim, L., Park, D.K.: Quantum convolutional neural network for classical data classification. Quantum Machine Intelligence 4(3) (2022) https://doi.org/ 10.1007/s42484-021-00061-x 204

  20. [26]

    Holmes, Z., Coble, N.J., Sornborger, A.T., Subasi, Y.: On nonlinear transfor- mations in quantum computation. Phys. Rev. Research 5(013105) (2023) https: //doi.org/10.1103/PhysRevResearch.5.013105

  21. [27]

    arXiv: 2204.00738 (2022) https: //doi.org/10.48550/arXiv.2204.00738

    Jing, H., Wang, Y., Li, Y.: Data-driven quantum approximate optimization algorithm for cyber-physical power systems. arXiv: 2204.00738 (2022) https: //doi.org/10.48550/arXiv.2204.00738

  22. [28]

    Journal of the London Mathematical Society 110(5) (2024)

    Junge, M., Palazuelos, C.: On the power of quantum entanglement in multipartite quantum xor games. Journal of the London Mathematical Society 110(5) (2024)

  23. [29]

    Physical Review A 103(052425) (2021) https://doi.org/10.1103/PhysRevA.103.052425

    Kubo, K., Nakagawa, Y.O., Endo, S., Nagayama, S.: Variational quantum simula- tions of stochastic differential equations. Physical Review A 103(052425) (2021) https://doi.org/10.1103/PhysRevA.103.052425

  24. [30]

    Linear Algebra and its Applications 400, 147–167 (2005) https://doi.org/10.48550/arXiv.math/ 0404553

    Kribs, D.W.: A quantum computing primer for operator theorists. Linear Algebra and its Applications 400, 147–167 (2005) https://doi.org/10.48550/arXiv.math/ 0404553

  25. [31]

    npj Quantum Information 4(14) (2008) https://doi.org/10.1038/s41534-018-0060-8

    Li, R.Y., Di Felice, R., Rohs, R., Lidar, D.A.: Quantum annealing versus classi- cal machine learning applied to a simplied computational biology problem. npj Quantum Information 4(14) (2008) https://doi.org/10.1038/s41534-018-0060-8

  26. [32]

    Quantum Information Processing 20(393) (2021) https://doi.org/10.1007/s11128-021-03331-6

    Mahdian, M., Yeganeh, H.D.: Toward a quantum computing algorithm to quan- tify classical and quantum correlation of system states. Quantum Information Processing 20(393) (2021) https://doi.org/10.1007/s11128-021-03331-6

  27. [33]

    Scientific Reports 12(6379) (2022) https://doi.org/10.1038/s41598-022-10339-0

    Maldonado, T.J., Flick, J., Krastanov, S., Galda, A.: Error rate reduction of single-qubit gates via noise-aware decomposition into native gates. Scientific Reports 12(6379) (2022) https://doi.org/10.1038/s41598-022-10339-0

  28. [34]

    Journal of Chemical Theory and Computation 8(8), 2564–2568 (2012) https://doi.org/10.1021/ct300544e

    Manby, F.R., Stella, M., Goodpaster, J.D., Miller, T.F.: A simple, exact density-functional-theory embedding scheme. Journal of Chemical Theory and Computation 8(8), 2564–2568 (2012) https://doi.org/10.1021/ct300544e

  29. [35]

    Mensa, S., Sahin, E., Tacchino, F., Barkoutsos, P.K., Tavernelli, I.: Quantum machine learning framework for virtual screening in drug discovery: a prospective quantum advantage. Mach. Learn.: Sci. Technol. 4(015023) (2023) https://doi. org/10.1088/2632-2153/acb900

  30. [36]

    Nan Sheng, H.M., Govono, M., Galli, G.: Quantum embedding theory for strongly-correlated states in materials. J. Chem. Theory Comput. 17(4), 2116– 2125 (2021) https://doi.org/10.1021/acs.jctc.0c01258

  31. [38]

    Quantum Information and Computation 16(13-14), 1191–1211 (2016) https://doi.org/10.26421/QIC16.13-14-6

    Ostrev, D.: The structure of nearly-optimal quantum strategies for the non-local xor games. Quantum Information and Computation 16(13-14), 1191–1211 (2016) https://doi.org/10.26421/QIC16.13-14-6

  32. [39]

    Physical Review A 107(032428) (2023) https://doi.org/10

    Paine, A.E., Elfving, V.E., Kyriienko, O.: Quantum kernel methods for solving differential equations. Physical Review A 107(032428) (2023) https://doi.org/10. 1103/PhysRevA.107.032428

  33. [40]

    Paudel, H.P., Syamlal, M., Crawford, S.E., Lee, Y.-L., Shugayev, R.A., Lu, P., Ohodnicki, P.R., Mollot, D., Duan, Y.: Quantum computing and simulations for energy applications: Review and perspective. ACS Eng. Au 3, 151–196 (2022) https://doi.org/10.1021/acsengineeringau.1c00033

  34. [41]

    Journal of Physics A: Mathematical and Theoretical 55(085301) (2022) https://doi.org/10.1088/1751-8121/ac4b15

    Przhiyalkovskiy, Y.V.: Quantum process in probability representation of quan- tum mechanics. Journal of Physics A: Mathematical and Theoretical 55(085301) (2022) https://doi.org/10.1088/1751-8121/ac4b15

  35. [42]

    Physics Reports 687, 1– 51 (2017) https://doi.org/https://papers.ssrn.com/sol3/papers.cfm?abstract id= 2972841

    Perc, M.: Statistical physics of human cooperation. Physics Reports 687, 1– 51 (2017) https://doi.org/https://papers.ssrn.com/sol3/papers.cfm?abstract id= 2972841

  36. [43]

    Ravishankar Ramanathan, R., Augusiak, R., Murta, G.: Generalized xor games withd outcomes and the task of nonlocal computation. Phys. Rev. A 93(022333) (2016) https://doi.org/10.1103/PhysRevA.93.022333

  37. [45]

    arXiv: 2209.07714 (2022) https://doi.org/10.48550/arXiv.2209

    Rigas, P.: Variational quantum algorithm for measurement extraction from the navier-stokes, einstein, maxwell, b-type, lin-tsien, camassa-holm, dsw, h-s, kdv-b, non-homogeneous kdv, generalized kdv, kdv, translational kdv, skdv, b-l and airy equations (v4). arXiv: 2209.07714 (...

  38. [46]

    Journal of Phys A: Math

    Roscika, M., Mazurek, P., Grudka, A., Horodecki, M.: Generalized xor non- locality games with graph description on a square lattice. Journal of Phys A: Math. Theor. 53(265302) (2020) https://doi.org/10.1088/1751-8121/ab8f3e

  39. [47]

    Journal of Mathematical Physics 52(10), 102202 (2011) https://doi.org/ 10.1063/1.3652924

    Slofstra, W.: Lower bounds on the entanglement needed to play xor non-local games. Journal of Mathematical Physics 52(10), 102202 (2011) https://doi.org/ 10.1063/1.3652924

  40. [48]

    Diversities in Quantum Computation and Quantum Information, 79–105 (2012) https://doi.org/10.1142/9789814425988 0003 206

    Dam, W., Sasaki, Y.: Quantum algorithms for problems in number theory, alge- braic geometry, and group theory. Diversities in Quantum Computation and Quantum Information, 79–105 (2012) https://doi.org/10.1142/9789814425988 0003 206

  41. [50]

    Wang, Y., Krstic, P.S.: Multistate transition dynamics by strong time-dependent perturbation in nisq era. J. Phys. Commun. 7(075004) (2023) https://doi.org/10. 1088/2399-6528/ace67a

  42. [51]

    Quantum Machine Intelligence 3(21) (2021) https://doi.org/10.1007/s42484-021-00048-8 207

    Zhao, L., Zhao, Z., Rebentrost, P., Fitzsimons, J.: Compiling basic linear algebra subroutines for quantum computers. Quantum Machine Intelligence 3(21) (2021) https://doi.org/10.1007/s42484-021-00048-8 207

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.