Certifying Majorana Fermions with Elegant-Like Bell Inequalities and a New Self-Testing Equivalence

Patryk Michalski, Arturo Konderak, Wojciech Bruzda, Remigiusz Augusiak · 2025

Read Certifying Majorana Fermions with Elegant-Like Bell Inequalities and a New Self-Testing Equivalence by Patryk Michalski, Arturo Konderak, Wojciech Bruzda, Remigiusz Augusiak (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2511-17764.txt.

Introduction

Patryk Michalski, Arturo Konderak, Wojciech Bruzda, and Remigiusz Augusiak build new Bell inequalities, tests that separate quantum links between particles from any classic local story. For most such tests the best possible quantum score is hard to find exactly, which limits their use.

Their construction gives that bound exactly and generalizes two classic inequalities. Under stated conditions the tests can certify Majorana fermions in a device independent way, meaning the claim rests on observed data alone. They also identify a new equivalence that must be part of self testing, the standard for certifying a quantum setup.

Catalog record

CREATORPatryk Michalski, Arturo Konderak, Wojciech Bruzda, Remigiusz Augusiak
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE80 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2511-17764

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Certifying Majorana Fermions with Elegant-Like Bell Inequalities and a New Self-Testing Equivalence Patryk Michalski, Arturo Konderak, Wojciech Bruzda, Remigiusz Augusiak Abstract Bell inequalities provide a fundamental tool for probing nonlocal correlations, yet their quantum bound, that is, the maximal value attainable through quantum strategies, is rarely accessible analytically. In this work, we introduce a general construction of Bell inequalities for which this bound can be computed exactly. Our framework generalizes both the Clauser–Horne–Shimony–Holt and Gisin’s elegant inequalities, yielding Bell expressions maximally violated by any number of pairwise anticommuting Clifford observables together with the corresponding maximally entangled state. Under suitable assumptions, our inequalities also enable the device-independent certification of Majorana fermions, understood as multiqubit realizations of Clifford algebra generators. Importantly, we identify an additional equivalence that must be incorporated into the definition of self-testing beyond invariance under local isometries and transposition. This equivalence arises from partial transposition applied to the shared state and to the measurements, which in specific cases leaves all observed correlations unchanged. I Introduction One of the central distinctions between classical and quantum physics lies in the structure of correlations between distant systems. In classical theories, such correlations can always be attributed to local hidden variable (LHV) models. Quantum mechanics, in contrast, predicts correlations that cannot be reproduced by any LHV description. This phenomenon, which is a consequence of quantum entanglement, reveals the intrinsically nonlocal character of quantum theory and is known as Bell nonlocality. To formalize this distinction, Bell inequalities were introduced as constraints that any LHV theory must satisfy, yet which can be violated by quantum mechanics 11 . Such violations not only reveal the presence of nonlocal correlations, but also constitute a powerful resource for tasks such as quantum cryptography 2 and quantum communication complexity 16 ; 17 . Moreover, the violation of suitably chosen Bell inequalities enables self-testing 43 ; 55 —the strongest form of device-independent certification—which allows one to infer both the underlying quantum state and the performed measurements solely from the observed correlations. The device-independent approach to quantum certification eliminates the need to trust the internal functioning of the devices, making it particularly relevant for secure quantum technologies. This paradigm has attracted significant attention in recent years, leading to numerous self-testing results for entangled quantum states. Notably, all pure bipartite entangled states 21 and all multiqubit entangled states 7 are now known to be self-testable. Parallel advances have established self-testing for various families of measurements, including Pauli observables 13 , all two-dimensional projective measurements 61 , certain d -outcome observables 52 , and constant-size measurements with algebraic structure 41 . More recent developments extend these results to all real projective measurements 18 and, in network scenarios, even to arbitrary quantum measurements 51 . Although Bell inequalities that enable self-testing are of great interest, designing them for specific states and measurements remains a notoriously challenging task (see, however, Ref. 9 ). Our work aims to partially fill this gap. We introduce a framework for constructing bipartite correlation Bell inequalities with arbitrary numbers of binary observables, based on a class of rectangular matrices with pairwise orthogonal rows and normalized columns. We refer to these as row-orthogonal, column-normalized (ROCN) matrices, and to the associated Bell expressions as ROCN Bell inequalities. A key feature of this family is that their quantum bound can be computed analytically; moreover, it is in general tighter than the best-known universal upper bound for correlation Bell inequalities derived in Ref. 27 . We identify conditions under which ROCN Bell inequalities guarantee a quantum advantage over all LHV models. Remarkably, these inequalities are maximally violated by maximally entangled states together with sets of Clifford observables of arbitrary size. In this sense, our construction extends the well-known Clauser–Horne–Shimony–Holt 19 and Gisin’s elegant 28 Bell inequalities to arbitrarily large families of binary, pairwise anticommuting observables. A further contribution of our work is the self-testing of Majorana operators (or fermions), which provide a concrete physical realization of Clifford algebra generators. Majorana operators play a central role across several areas of physics: they arise in quantum field theory 31 and quantum statistical mechanics 5 (see also Ref. 15 ). In quantum computation, they are particularly important due to their connection to the Pauli group and the stabilizer formalism 45 . For an overview of their broader role in fermionic systems, see Ref. 56 . They also feature prominently in entropic uncertainty relations, achieving maximal violations in scenarios involving more than two measurements 60 ; 22 . A contextuality-based approach for certifying two-qubit Majorana fermions was previously proposed in Ref. 36 . Importantly, for an odd number of Majorana operators, we show that self-testing is defined only up to a specific transformation that generalizes the well-known invariance of quantum correlations under complex conjugation in the case of three Pauli matrices. This reveals a previously unrecognized form of equivalence that should be incorporated into the definition of self-testing. In fact, it turns out that apart from local isometries and complex conjugation (cf. Ref. 55 ), in certain situations, the observed correlations can also be invariant under partial transposition of the state and the measurement. II Preliminaries Before getting to the results, let us first establish the terminology and notation that will be consistently employed throughout the manuscript, and recall the most relevant concepts and facts concerning Bell nonlocality and self-testing. II.1 Bell nonlocality & Bell inequalities We consider the standard Bell scenario involving two spacelike separated observers, Alice and Bob, who share a quantum state \rho_{\mathrm{AB}} that acts on some finite-dimensional composite Hilbert space \mathcal{H}_{\mathrm{AB}}=\mathcal{H}_{\mathrm{A}}\otimes\mathcal{H}_{\mathrm{B}} . On their respective subsystems, Alice and Bob perform measurements chosen from sets of m and n independent two-outcome observables, denoted respectively as A_{i} and B_{j} , with i\in\{1,\ldots,m\} and j\in\{1,\ldots,n\} . Formally, A_{i} and B_{j} are Hermitian operators acting on \mathcal{H}_{\mathrm{A}} and \mathcal{H}_{\mathrm{B}} , respectively, such that A_{i}^{2}=B_{j}^{2}=\mathbbm{1} (where, with a slight abuse of notation, \mathbbm{1} denotes the identity on the corresponding local Hilbert space). Observe that, due to Naimark’s dilation theorem 33 , it is sufficient to consider projective measurements 8 . The outcomes observed by Alice and Bob are denoted a and b , where a,b=\pm 1 . The correlations generated when the parties perform many rounds of measurements are described by a collection of probability distributions \{p(a,b|i,j)\} , where p(a,b|i,j) denotes the probability that Alice and Bob obtain outcomes a and b upon performing measurements i and j . Since we deal here with two-outcome measurements, one can equivalently represent the observed correlations through a collection of expectation values 29 : \vec{c}\equiv\{\langle A_{i}B_{j}\rangle,\langle A_{i}\rangle,\langle B_{j}\rangle\}, (1) which are given by Born rule as \displaystyle\langle A_{i}B_{j}\rangle=\Tr[\rho_{\mathrm{AB}}(A_{i}\otimes B_{j})], \displaystyle\langle A_{i}\rangle=\Tr[\rho_{\mathrm{AB}}(A_{i}\otimes\mathbbm{1})],\quad\langle B_{j}\rangle=\Tr[\rho_{\mathrm{AB}}(\mathbbm{1}\otimes B_{j})]. Both collections (of probability distributions or expectation values) are commonly referred to as correlations or behaviour . Let us now consider correlations for which the joint expectation values admit a local hidden-variable (LHV) decomposition of the form \langle A_{i}B_{j}\rangle=\sum_{\lambda}p(\lambda)\,\langle A_{i}\rangle_{\lambda}\langle B_{j}\rangle_{\lambda}, (3) where \lambda denotes the hidden variable (or shared randomness), p(\lambda) is a probability distribution, and -1\leqslant\langle A_{i}\rangle_{\lambda},\langle B_{j}\rangle_{\lambda}\leqslant 1 for any \lambda and any pair i,j . The set of all correlations admitting such a decomposition forms a convex polytope, whose vertices correspond to local deterministic correlations satisfying \langle A_{i}B_{j}\rangle=\langle A_{i}\rangle\langle B_{j}\rangle for every pair i,j ,

Read the complete text