Pith. sign in

REVIEW 3 major objections 6 minor 2 cited by

Two families of Bell inequalities are constructed so that, under a uniqueness assumption on the maximal quantum violation, their outcomes certify 2 log d bits of global randomness for any dimension d.

Reviewed by Pith at T0; open to challenge. T0 means a machine referee read the full paper against a public rubric. the ladder, T0–T4 →

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

T0 review · deepseek-v4-flash

2026-08-02 10:37 UTC pith:33ZB2RN5

load-bearing objection Two explicit Bell families with a genuine d=3 self-test; the arbitrary-d randomness claim is a clearly labelled conjecture, not a theorem. the 3 major comments →

arxiv 2606.21362 v3 pith:33ZB2RN5 submitted 2026-06-19 quant-ph

Bell inequalities tailored to optimal global randomness certification

Ignacio Perito , Raffaele D'Avino , Micha{\l} Jung , Piotr Mironowicz , Antonio Ac\'in , Remigiusz Augusiak This is my paper
classification quant-ph MSC 81P4081P45 PACS 03.67.-a03.65.Ud03.67.Mn
keywords Bell inequalitiesdevice-independent randomnessglobal randomness certificationquditsmaximally entangled statesself-testingsymmetry argumentsquantum nonlocality
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved

The pith

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

This paper aims to show that, for bipartite Bell tests with d-outcome measurements, there exist inequalities whose maximal quantum violation forces the joint outputs of a chosen pair of measurements to be perfectly uniform over all d² possibilities, giving 2 log d bits of certified randomness. For the first family, the authors use a symmetry argument and conjecture the optimal violation, supporting it numerically for d=3 and d=4. For the second family, they prove the quantum bound analytically for every d and, in the d=3 case, provide a full self-testing proof that the violation certifies two fully random trits. The main caveat, which the paper states explicitly, is that uniqueness of the maximal violation is not proven for general d, so the global-randomness conclusion for arbitrary d rests on that assumption.

Core claim

The central claim is that maximal quantum violations of the two constructed Bell functionals certify that Alice and Bob's outputs for a designated pair of settings are uniformly distributed over all d² joint outcomes. The mechanism is a symmetry argument: the functional is invariant under transformations that multiply certain phase correlators by nontrivial roots of unity; if the maximal violation is unique up to symmetries, those correlators must vanish, and Fourier inversion then forces p(ab|xy)=1/d². For the first family, the conjectured maximal value is 2/sin(pi/2d)+1, with the certification verified numerically for d=3,4. For the second family, the paper proves a quantum bound of exactl

What carries the argument

The central object is a Bell functional expressed through Fourier correlators ⟨A_x^(k) B_y^(l)⟩, where ω is a primitive d-th root of unity. Symmetry transformations T and T′ leave the functional invariant while multiplying designated correlators by nontrivial phases; combined with uniqueness of the maximal violation, this forces all nontrivial correlators for the target settings to vanish, which is equivalent to a flat joint output distribution. The explicit optimal strategies use a maximally entangled state of two qudits and generalized clock-and-shift Pauli operators, with Bob's observables chosen through specific coefficients so that the relevant operator sums stabilize the maximally enta

Load-bearing premise

The claim that the maximal violations certify uniform joint randomness relies on the unproven assumption that, for every dimension d, the quantum state and measurements achieving the maximal violation are unique up to the obvious symmetries.

What would settle it

Find, for some d≥4, a quantum strategy that achieves the conjectured maximal violation of the first family's Bell functional but gives a non-uniform joint distribution for the designated settings (x=1, y=d); equivalently, exhibit two inequivalent optimal strategies for the same violation. Either would break the symmetry argument and refute the global-randomness claim for that dimension. For the second family, the analogous test is to find d≥4 optimal strategies for the extended functional that are not of the constructed form.

Watch this falsifier — get emailed when new claim-graph text bears on it.

If this is right

  • If the conjectures hold, a single Bell test on d-outcome devices can certify the maximum possible global randomness, 2 log d bits, against any adversary correlated with the devices.
  • For the second family, the analytic quantum bound and the d=3 self-testing result give a rigorous template for device-independent randomness certification in higher dimensions once uniqueness is established.
  • The first family uses only two measurement settings on Alice's side, making it comparatively practical; for d=3 it tolerates up to 25% depolarizing noise, which the paper notes is better than previously known inequalities in the same task.
  • The symmetry method, originally developed for binary outcomes, is extended here to arbitrary outcome number, providing a general design principle for Bell inequalities tailored to global randomness.

Where Pith is reading between the lines

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

  • The symmetry argument likely generalizes beyond these two families: any Bell functional invariant under phase shifts of a designated correlator, with a unique maximizer, will certify full joint uniformity, so this could become a standard recipe for constructing randomness-certifying inequalities.
  • The sum-of-squares decomposition for the second family may lead to self-testing of generalized X and Z observables in every dimension, thereby providing device-independent certification of mutually unbiased bases; the paper mentions this as an open problem, but it is a natural next step.
  • A concrete experiment on a qudit photonic or trapped-ion platform could test the d=3 protocol; measuring the certified randomness under controlled depolarizing noise would verify the predicted noise robustness.
  • Because the paper notes that concurrent independent constructions exist, a direct comparison of measurement overhead and noise tolerance between the two approaches would be a useful practical follow-up.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

3 major / 6 minor

Summary. The paper constructs two families of bipartite Bell functionals, \bar I_d and \bar F_d, aimed at certifying 2\log d bits of global randomness in a device-independent manner. For the first family, defined in the (2,d+1;d,d) scenario, the authors derive the classical and non-signaling bounds for the reduced functional I_d, give an analytical upper bound d\sqrt{2}, and exhibit an explicit quantum strategy attaining 2(\sin(\pi/2d))^{-1}, which they conjecture to be the Tsirelson bound. They support this with NPA computations up to d=6 and with numerical guessing-probability bounds for d=3,4. For the second family F_d, defined in the (d,d;d,d) scenario, they prove via an SOS decomposition that the Tsirelson bound is exactly d for all d, and give an explicit optimal realization. For d=3 they prove a full self-testing theorem, which implies certification of two fully random trits from the maximal violation of \bar F_3. The arbitrary-dimensional global-randomness statements are explicitly conditional on unproven uniqueness of the maximizers and, for I_d, on Conjecture 1.

Significance. If the conjectures hold, the paper would provide the first systematic construction of Bell inequalities tailored to optimal global randomness certification with an arbitrary number of outputs. The first family is experimentally appealing because Alice needs only two measurements; the second family has an analytically proven Tsirelson bound for all d and a complete self-test for d=3. The methods are clean: the symmetry argument is standard, and the SOS decomposition for F_d is elegant. The paper is transparent about its conjectural steps and gives machine-independent, parameter-free derivations of the classical, non-signaling, and (for F_d) quantum bounds. The main value is the explicit families and the reduction of arbitrary-d randomness certification to a clearly stated uniqueness problem.

major comments (3)
  1. [Sec. III.B, Conjectures 1 and 2; Sec. IV, uniqueness paragraph; Sec. V] The headline claim of optimal global randomness certification for arbitrary d is not proven. For the first family, Conjecture 1 is verified numerically only up to d=6 (Table I), and Conjecture 2 only for d=3,4 (Table II). For the second family, Theorem 1 proves the Tsirelson bound for all d, but the step from maximal violation to p(ab|1,d)=1/d^2 requires uniqueness of the maximizer, which the text explicitly states is unproven: 'if this maximal violation happens to be unique. This is yet unproven but seems plausible' (Sec. IV). Nevertheless, Sec. V states that the families 'allow to certify optimal global randomness using maximally entangled states of arbitrary local dimension.' This overstates the proven content. Please either supply a proof of uniqueness or a direct guessing-probability bound for d>3, or systematically rephrase all global claims as conditional on Conjectures 1-2 and th
  2. [Sec. IV and Appendix D] For \bar F_d with d>3, no randomness or self-testing statement is established. The general SOS (22) at maximal violation gives P_\ell|\psi\rangle=0, but the d=3 proof (Appendix D) uses the stronger equations L_i|\psi\rangle=L_i^\dagger|\psi\rangle=0 and the classification of solutions from Ref. [31]. For d>3, P_\ell^\dagger|\psi\rangle=0 is not obtained, so the uniqueness assumption behind the symmetry argument is not supported by any structural result. Moreover, no numerical guessing-probability computation for \bar F_d is reported for any d>3. I ask the authors to either add such evidence (e.g., NPA-based guessing bounds for d=4,5) or explicitly label the d>3 global-randomness claim of the second family as a conjecture, separate from the proven d=3 result.
  3. [Sec. III.A and Table II] The numerical claims of 'certification' for d=3,4 are based on level 1+AB of the NPA hierarchy. The tabulated upper bounds (0.11112 vs 1/9 and 0.06251 vs 1/16) are outer-approximation bounds at a fixed relaxation level, not rigorous device-independent certificates. The text says 'up to numerical precision,' but the abstract states the d=3 result without this caveat. Please clarify in the main text that these are numerical SDP bounds with a specified tolerance, and that tightness of the relaxation for the guessing-probability problem is not proven.
minor comments (6)
  1. [General notation] The Bell functional is introduced as \bar I_d, but Appendix B and Table II use \tilde I_d without definition. Unify the notation.
  2. [Sec. III.A] 'two unbiased bases' should be 'two mutually unbiased bases' (MUBs), to match the standard terminology used later.
  3. [Appendix C] The operation 'smodd' is used without a prior definition. Define it explicitly before Eq. (C16), and consider giving a simpler closed form for the phases.
  4. [Theorem 2] The statement 'attained uniquely (up to local isometries and transposition map)' is stronger than what the proof shows: the observables are determined up to a direct sum of the reference representation and its transpose on auxiliary subspaces (P_1+P_2=1, Q_1+Q_2=1). Rephrase to reflect the direct-sum freedom.
  5. [Table I] The table would be easier to read if the NPA level at which each numerical bound is tight were explicitly listed, and if the dashes were explained (e.g., not computed at that level).
  6. [Abstract and text] The abstract's '2\log d random bits' is ambiguous about the logarithm base; use \log_2 d consistently. Also fix the spelling 'Krönecker' to 'Kronecker' in Sec. III.B.

Circularity Check

0 steps flagged

No significant circularity: the derivations are explicit constructions, SDP/NPA bounds, and conditional symmetry arguments; the unproven uniqueness assumptions are acknowledged limitations, not circular inputs.

full rationale

The paper's derivation chain is not circular. The first family's quantum value is obtained by an explicit construction (Eqs. 13-15 and Appendix C), with NPA upper bounds and see-saw lower bounds used to support Conjecture 1. The optimal-global-randomness certification for d=3,4 is obtained by a device-independent SDP on the guessing probability under the Bell-violation constraint, not by assuming the certified uniformity. The symmetry argument in Sec. II is a conditional implication: if the maximal-violating correlation is unique and the Bell functional is invariant under T and T', then off-diagonal Fourier correlators vanish; the paper explicitly states that uniqueness is unproven for I3 and for Fd beyond d=3 ('While we cannot prove analytically the uniqueness of the maximal quantum violation of I3', 'This is yet unproven but seems plausible'). This is an openly acknowledged assumption/limitation, not a circular reuse of the target conclusion. The d=3 self-test (Theorem 2, Appendix D) reduces the maximal-violation equations to a system already solved in Ref. [31]; although Ref. [31] shares an author, it is an external published classification of two-qutrit observables whose assumptions do not include the present Bell functional, so it provides legitimate mathematical support rather than circular self-citation. The Tsirelson bound for Fd (Theorem 1) follows from an explicit SOS decomposition (Eq. 22) and the normalization of the λ coefficients, independent of the randomness claim. No fitted parameters are relabeled as predictions, and no input is defined in terms of the output. The acknowledged gaps are about unproven uniqueness and missing analytical proofs for arbitrary d, which are correctness/completeness risks, not circularity.

Axiom & Free-Parameter Ledger

0 free parameters · 5 axioms · 0 invented entities

The central claims do not use fitted parameters; all coefficients are explicit. However, the conclusions for arbitrary d rest on three unproven premises: the uniqueness of the maximal violation (needed for the symmetry argument), the optimality of the constructed strategy for I_d (Conjecture 1), and the optimal-randomness conjecture for \bar{I}_d (Conjecture 2). For F_d, the Tsirelson bound is proven, but the uniqueness needed for global randomness is only self-tested for d=3. These are ad hoc to this paper's program.

axioms (5)
  • ad hoc to paper The maximal quantum violation of \bar{I}_d is unique (up to local symmetries) for all d
    Required for the symmetry argument to force zero off-diagonal correlators and hence uniform randomness; explicitly stated as unproven in Sec III.A for d=3 and Sec III.B generally.
  • ad hoc to paper The constructed realization attains the Tsirelson bound of I_d for all d (Conjecture 1)
    The quantum violation 2/sin(π/2d) is used to set the certificate level; verified numerically only for d≤6 (Appendix A3).
  • ad hoc to paper The maximal violation of \bar{F}_d is unique for all d, implying global randomness from the symmetry argument
    Explicitly stated as unproven in Sec IV; self-testing is established only for d=3 (Theorem 2).
  • domain assumption Quantum theory is a valid and complete description of the Bell setup, with projective d-outcome measurements
    Standard Bell-scenario assumptions stated in Sec II, Eq. (1).
  • standard math NPA hierarchy upper bounds and see-saw lower bounds faithfully approximate the quantum set
    Standard SDP tools used in Appendix B; numerical results rely on these relaxations.

pith-pipeline@v1.3.0-alltime-deepseek · 17604 in / 13582 out tokens · 115627 ms · 2026-08-02T10:37:37.933144+00:00 · methodology

0 comments
read the original abstract

We present two novel families of bipartite Bell inequalities designed to achieve optimal global randomness certification for an arbitrary number of outputs $d$. We first use symmetry arguments to argue that their maximal quantum violations certify $2\log d$ random bits. For the first family, we construct a quantum realization using $d\times d$ maximally entangled states which provides a quantum violation that we conjecture to be optimal for any $d$. It is then numerically shown that the obtained quantum violation certifies optimal global randomness, up to numerical precision, for $d=3,4$. For the second family, we provide the optimal quantum violation and its quantum realization for any $d$, again using $d\times d$ maximally entangled states and projective measurements over at least two unbiased bases on one of the parties. We self-test this realization for $d=3$, which implies the optimal certification of two fully random trits.

Figures

Figures reproduced from arXiv: 2606.21362 by Antonio Ac\'in, Ignacio Perito, Micha{\l} Jung, Piotr Mironowicz, Raffaele D'Avino, Remigiusz Augusiak.

Figure 1
Figure 1. Figure 1: FIG. 1. Ratio between the local and quantum bounds of [PITH_FULL_IMAGE:figures/full_fig_p010_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: FIG. 2. Certified randomness as a function of the noise parameter [PITH_FULL_IMAGE:figures/full_fig_p012_2.png] view at source ↗

Review history (2 revisions) →

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

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

  1. Noise robustness of three outcome Bell certified quantum randomness

    quant-ph 2026-06 unverdicted novelty 6.0

    Numerical study finds many three-outcome Bell inequalities certify substantial min-entropy with noise robustness, including simple ones using fewer settings that achieve near-maximal randomness.

  2. Maximal global device-independent randomness from projective measurements in every dimension

    quant-ph 2026-06 unverdicted novelty 6.0

    Projective measurements using mutually unbiased bases achieve the theoretical maximum device-independent randomness rate of 2 log(d) bits from bipartite systems in every dimension d ≥ 2.

Reference graph

Works this paper leans on

43 extracted references · 2 linked inside Pith · cited by 2 Pith papers

  1. [1]

    It turns out that such a choice of Bob’s measurements is possible and given by: By = 1 3 2Z 2 + 2ω2yX−ω y+1X 2Z (9) (see Appendix C for details)

    by bothZ⊗Z 2 andX⊗X, we have that this strategy will saturate the quantum bound⟨I3⟩= 4iffα= 2. It turns out that such a choice of Bob’s measurements is possible and given by: By = 1 3 2Z 2 + 2ω2yX−ω y+1X 2Z (9) (see Appendix C for details). From here, and given that the state is maximally entangled, it is easy to find an optimal strategy for ¯I3: we take ...

    2021

  2. [2]

    Li and P

    M. Li and P. M. B. Vitanyi,An introduction to kol- mogorov complexity and its applications, 3rd ed., Texts in Computer Science (Springer, New York, NY, 2008)

    2008

  3. [3]

    Then, we choose Bob’s measurements in such a way that: B0 +B 1 +B 2 =αZ 2, B 0 +ωB 1 +ω 2B2 =αX,(8) whereαis a constant that we aim to maximize

    P k |kk⟩, and Alice’s measurements as the gen- eralizedZandXoperators: A0 =Z= 2X k=0 ωk |k⟩ ⟨k|, A1 =X= 2X k=0 |k+ 1⟩ ⟨k|.(7) Note that they give two unbiased bases. Then, we choose Bob’s measurements in such a way that: B0 +B 1 +B 2 =αZ 2, B 0 +ωB 1 +ω 2B2 =αX,(8) whereαis a constant that we aim to maximize. As the state|ψ 3⟩is stabilized (it is an eigen...

  4. [4]

    J. S. Bell, On the Einstein Podolsky Rosen paradox, Physics Physique Fizika1, 195 (1964)

    1964

  5. [5]

    Acín and L

    A. Acín and L. Masanes, Certified randomness in quan- tum physics, Nature540, 213 (2016)

    2016

  6. [6]

    J. F. Clauser, M. A. Horne, A. Shimony, and R. A. Holt, Proposed experiment to test local hidden-variable theo- ries, Phys. Rev. Lett.23, 880 (1969)

    1969

  7. [7]

    A. Acín, S. Massar, and S. Pironio, Randomness ver- sus nonlocality and entanglement, Phys. Rev. Lett.108, 100402 (2012)

    2012

  8. [8]

    Ringbauer, M

    M. Ringbauer, M. Meth, L. Postler, R. Stricker, R. Blatt, P. Schindler, and T. Monz, A universal qudit quantum processor with trapped ions, Nat. Phys.18, 1053 (2022)

    2022

  9. [9]

    Y. Chi, Y. Yu, Q. Gong, and J. Wang, High-dimensional quantum information processing on programmable inte- grated photonic chips, Science China Information Sci- ences66, 180501 (2023)

    2023

  10. [10]

    Sarkar, D

    S. Sarkar, D. Saha, J. Kaniewski, and R. Augusiak, Self- testing quantum systems of arbitrary local dimension with minimal number of measurements, npj Quantum Information7, 151 (2021)

    2021

  11. [11]

    J. J. Borkała, C. Jebarathinam, S. Sarkar, and R. Au- gusiak, Device-independent certification of maximal ran- domness from pure entangled two-qutrit states using non-projective measurements, Entropy (Basel)24, 350 (2022)

    2022

  12. [12]

    Farkas, J

    M. Farkas, J. Volčič, S. A. L. Storgaard, R. Chen, and L. Mančinska, Maximal device-independent randomness in every dimension, Nature Physics22, 319 (2026)

    2026

  13. [13]

    For simplicity, in this work we usedto denote the number of outcomes in the Bell test and the dimension of the considered quantum systems

  14. [14]

    Mironowicz and M

    P. Mironowicz and M. Pawłowski, Robustness of quantum-randomness expansion protocols in the pres- ence of noise, Physical Review A88, 032319 (2013)

    2013

  15. [15]

    Y. Z. Law, L. P. Thinh, J.-D. Bancal, and V. Scarani, Quantum randomness extraction for various levels of characterization of the devices, Journal of Physics A: Mathematical and Theoretical47, 424028 (2014)

    2014

  16. [16]

    Woodhead, J

    E. Woodhead, J. Kaniewski, B. Bourdoncle, A. Salavrakos, J. Bowles, A. Acín, and R. Augu- siak, Maximal randomness from partially entangled states, Phys. Rev. Res.2, 042028(R) (2020)

    2020

  17. [17]

    de la Torre, M

    G. de la Torre, M. J. Hoban, C. Dhara, G. Prettico, and A. Acín, Maximally nonlocal theories cannot be maxi- mally random, Phys. Rev. Lett.114, 160502 (2015)

    2015

  18. [18]

    Dhara, G

    C. Dhara, G. Prettico, and A. Acín, Maximal quantum randomnessinBelltests,Phys.Rev.A88,052116(2013)

    2013

  19. [19]

    Navascués, S

    M. Navascués, S. Pironio, and A. Acín, Bounding the set of quantum correlations, Phys. Rev. Lett.98, 010401 (2007)

    2007

  20. [20]

    Navascués, S

    M. Navascués, S. Pironio, and A. Acín, A convergent hi- erarchy of semidefinite programs characterizing the set of quantum correlations, New Journal of Physics10, 073013 (2008)

    2008

  21. [21]

    D’Avino, I

    R. D’Avino, I. Perito, P. Mironowicz, A. Acín, and R. Au- gusiak, Noise robustness of three outcome bell certified quantum randomness (2026), arXiv:2606.21371 [quant- ph]. 7

    Pith/arXiv arXiv 2026

  22. [22]

    Without losing generality, the number of outcomes can be taken equal for all measurements by Alice and Bob, as if some measurements have less outputs than others, one can always pickdas the maximum and assign a zero probability to some outcomes

  23. [23]

    Zhang, Y

    C. Zhang, Y. Li, X.-M. Hu, Y. Xiang, C.-F. Li, G.-C. Guo, J. Tura, Q. Gong, Q. He, and B.-H. Liu, Random- ness versus nonlocality in multiple-input and multiple- output quantum scenario, Physical Review Letters134, 090201 (2025)

    2025

  24. [24]

    Nieto-Silleras, S

    O. Nieto-Silleras, S. Pironio, and J. Silman, Using complete measurement statistics for optimal device- independent randomness evaluation, New Journal of Physics16, 013035 (2014)

    2014

  25. [25]

    Bancal, L

    J.-D. Bancal, L. Sheridan, and V. Scarani, More ran- domness from the same data, New Journal of Physics 16, 033011 (2014)

    2014

  26. [26]

    Mironowicz, Semi-definite programming and quantum information, Journal of Physics A: Mathematical and Theoretical57, 163002 (2024)

    P. Mironowicz, Semi-definite programming and quantum information, Journal of Physics A: Mathematical and Theoretical57, 163002 (2024)

    2024

  27. [27]

    Gallego, N

    R. Gallego, N. Brunner, C. Hadley, and A. Acín, Device- independent tests of classical and quantum dimensions, Phys. Rev. Lett.105, 230501 (2010)

    2010

  28. [28]

    B. S. Cirel’son, Quantum generalizations of bell’s in- equality, Letters in Mathematical Physics4, 93 (1980)

    1980

  29. [29]

    McKague, T

    M. McKague, T. H. Yang, and V. Scarani, Robust self- testing of the singlet, Journal of Physics A: Mathematical and Theoretical45, 455304 (2012)

    2012

  30. [30]

    Buhrman and S

    H. Buhrman and S. Massar, Causality and Tsirelson’s bounds, Phys. Rev. A72, 052103 (2005)

    2005

  31. [31]

    R. F. Werner and M. M. Wolf, Bell inequalities and en- tanglement, Quantum Info. Comput.1, 1–25 (2001)

    2001

  32. [32]

    K. F. Pál and T. Vértesi, Maximal violation of a bipartite three-setting, two-outcome Bell inequality using infinite- dimensional quantum systems, Phys. Rev. A82, 022116 (2010)

    2010

  33. [33]

    Šupić, J.Tura, F.Baccari, A.Salavrakos, and R

    J.Kaniewski, I. Šupić, J.Tura, F.Baccari, A.Salavrakos, and R. Augusiak, Maximal nonlocality from maximal en- tanglement and mutually unbiased bases, and self-testing of two-qutrit quantum systems, Quantum3, 198 (2019)

    2019

  34. [34]

    Farkas, P

    M. Farkas, P. Mironowicz, and R. Augusiak, Maxi- mal global device-independent randomness from pro- jective measurements in every dimension (2026), arXiv:2606.21369 [quant-ph]

    Pith/arXiv arXiv 2026

  35. [35]

    A. C. Doherty, Y.-C. Liang, B. Toner, and S. Wehner, The quantum moment problem and bounds on entan- gledmulti-provergames,in2008 23rd Annual IEEE Con- ference on Computational Complexity(IEEE, 2008) pp. 199–210. 8 Appendix A: Relevant bounds forId

    2008

  36. [36]

    Optimal classical and non-signaling values forId Proposition 2.The non-signaling bound ofI d is: βN S(d) = 2d,(A1) and the classical bound is given by: βL(d) = 2  1 + 2 ⌊ d+1 2 ⌋−1X ℓ=1 cos πℓ d fℓ   ,(A2) wheref ℓ = cos(π/d)ifℓis odd andf ℓ = 1otherwise. Proof.The Bell operator can be written as: Id ≡ 1 2 d−1X y=0 (A0 +ω yA1)⊗B † y +h.c., (A3) and th...

  37. [37]

    Upper bound for the optimal quantum value ofId Proposition 3.The quantum bound ofI d satisfies: βQ(d)≤d √ 2.(A8) This upper bound is, in general, not tight. Proof.We define the operators: Pℓ ≡1− A0 +ω ℓA1√ 2 ⊗B ℓ,(A9) which satisfy: P † ℓ Pℓ =1− A0 +ω ℓA1√ 2 ⊗B ℓ + ω−ℓA† 0A1 2 ⊗1+h.c.,(A10) and hence: d−1X ℓ=0 P † ℓ Pℓ =d1− 1√ 2 d−1X ℓ=0 A0 +ω ℓA1 ⊗B ℓ + ...

  38. [38]

    Note that the upper bound (A8) is tight for level1of NPA in all cases, which is expected as the SOS decomposition used to prove it only involves level1terms

    Numerical bounds forI d The following table summarizes, ford≤6, numerical upper bounds (found by means of NPA relaxations) and lower bounds (by optimizing over explicit quantum strategies using a see-saw algorithm) forId and compares them to the analytical upper bound (A8) and the conjectured (and also analytical) optimal quantum value (17). Note that the...

  39. [39]

    For low dimensions, our construction exhibits a wide range of nonlocal values: it allows to certify nonlocality in presence of up to5%of depolarizing noise up tod= 9

    Ratio between local and quantum optimal values forId Figure 1 shows the ratioβL(d)/βQ(d)ford≤15. For low dimensions, our construction exhibits a wide range of nonlocal values: it allows to certify nonlocality in presence of up to5%of depolarizing noise up tod= 9. In particular, ford= 3, nonlocality can be witnessed even against25%of depolarizing noise. 10...

  40. [40]

    d Targetp guess Target randomnessUpper bound forp guess Lower bound for randomness 3 1/9 = 0.111

    Numerical proof of optimal randomness certification using˜Id ford= 3andd= 4 The following table summarizes numerical bounds for the device-independent guessing probability of both outputs when the full distributionp(ab|xy)is fixed by the optimal quantum strategy defined in Conjecture 1. d Targetp guess Target randomnessUpper bound forp guess Lower bound f...

  41. [41]

    (5), as well as˜F3

    Randomness robustness for ˜I3 and ˜F3 We analyze the robustness to noise of the inequality˜I3 defined in Eq. (5), as well as˜F3. The latter is obtained from the family ˜Fd, is defined as a modification ofFd (Eq. (19)) in the same spirit as for˜I3: namely, we include an additional measurement setting for Bob and introduce the extra correlator⟨A0B† d⟩. The ...

  42. [42]

    In [19], we perform a more systematic analysis of robustness, in which ˜I3 is compared with other known Bell inequalities with the same number of outcomes

    We observe that the robustness to noise of the two inequalities is very similar. In [19], we perform a more systematic analysis of robustness, in which ˜I3 is compared with other known Bell inequalities with the same number of outcomes. We find that, among inequalities that certify maximal randomness in the ideal (noise-free) scenario,˜I3 exhibits improve...

  43. [43]

    , d−1} ) .(C20) We use the same Dirichlet kernel as above, while shifting summation bounds asd−1 2 is not an integer

    The spectrum σ(Wy) = ( wy,j = d−1X k=0 (−1)kω−y(1+k) dsin π d (k+ 1 2 ) ω(j+ 3 2 )k, j∈ {0,1, . . . , d−1} ) .(C20) We use the same Dirichlet kernel as above, while shifting summation bounds asd−1 2 is not an integer. D d−1 2 π d (2k+ 1) = (−1)k sin π 2d (2k+ 1) = d−1 2X m=− d−1 2 eim( π d (2k+1)) = d 2 −1X m=− d 2 ei(m+ 1 2 )( π d (2k+1)).(C21) Now we re...