arXiv:2609.39411
Genuine Multipartite Nonlocality Is Fermionic Magic
Jamal Slim
Abstract
Free fermion dynamics is classically simulable, and a single non-Gaussian state makes it universal. We show that the same boundary governs quantum nonlocality. Measured as fermionic qubits, fermionic Gaussian states reach at most the biseparable bound of the Mermin inequality for three parties and, for any number of parties, violate the Mermin family by exactly the factor of a single entangled pair. They are never genuinely multipartite nonlocal. Genuine multipartite nonlocality and computational universality are thus one resource for fermions, the opposite of the stabilizer case. The mechanism is fermion parity. The bound is monogamous and certifies non-Gaussianity from measured correlations alone. On superconducting processors we watch nonlocality beyond free fermions switch on under an interaction between densities, confirm the monogamy law on 96 qubits, and measure the ground state of twelve interacting fermions on a ring cross the bound as the interaction grows, detecting the charge density wave transition.
AI Summary
The paper asks how much multipartite Bell nonlocality fermionic Gaussian states can support under local fermionic readout, and whether violations can witness non-Gaussianity, the resource enabling universal computation with matchgates. For two Majorana-bilinear measurements per party whose planes share a direction, it proves the tight bound $|\langle\prod_{j=1}^n(O_j^0+iO_j^1)\rangle|\le2^{n/2}$ for arbitrary pure or mixed Gaussian states. This covers arbitrary logical measurements of dual-rail and Majorana-tetron qubits. At three parties the bound is $2\sqrt2$, the biseparable Mermin bound, tightening to $2$ for anticommuting settings. It also excludes violations of the corresponding Svetlichny inequalities. The three-party optimum is resolved as a function of measurement angles. Proofs use Wick/Pfaffian identities, covariance contraction, and parity-dependent occupation-probability bounds. Gaussianity under occupation conditioning yields a parallel-repetition bound: $m$ disjoint Mermin games have joint success probability at most $[(2+\sqrt2)/4]^m$, even with correlations between cells. Superconducting-processor demonstrations report raw three-party Mermin values up to $3.538$, four-party values up to $7.02$ against a Gaussian bound of $4$, and sixteen-game success $0.174$ against $0.079$ on 96 qubits. A twelve-site interacting-ring simulation crosses the six-party Gaussian bound $8$ as interactions increase. The nonlocality conclusions concern the specified measurements and inequality families, not arbitrary Bell tests. For unrestricted bilinear settings, only a three-party bound of $3.26$ is certified; $2\sqrt2$ remains conjectural. Other open issues include thermodynamic transition-witness scaling and quantitative links to classical simulation cost. Hardware tests are nonadversarial benchmarks, not loophole-free Bell tests.
Central claim may not hold: The paper’s bounds on the Mermin family and specified Svetlichny expressions do not establish its categorical claim that Gaussian states cannot exhibit genuine multipartite nonlocality, because those inequalities are not shown to characterize all hybrid-local correlations. The abstract’s claim that a non-Gaussian state makes free fermion computation universal omits the essential pure-state qualification; non-Gaussian convex mixtures of Gaussian states provide counterexamples to the unrestricted resource identification.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.