arXiv:2610.06755
The Grothendieck Game and Self-Testing Continuous Groups
Alexander Kulpe, Giulio Malavolta, Simon Schmidt, Michael Walter
Abstract
Self-testing certifies quantum states and measurements from observed correlations alone, without requiring a description of the devices that produce them. In this work we establish, to the best of our knowledge, the first robust self-test of the full continuous family of measurements generated by Majorana operators. We achieve this by showing that the Grothendieck game, a two-player nonlocal game in which the players receive directions on a sphere and return binary answers, is a self-test of Majorana observables. That is, any near-optimal performance certifies an entire continuously parameterized family of quantum measurements. We further extend this certification to products of Majorana observables, obtaining a robust self-test of the spin representation of the continuous pin group.
AI Summary
Can a single nonlocal game robustly self-test a continuously indexed measurement family? This paper studies the Grothendieck XOR game $G_n$: questions are sphere directions $(x,y)$ sampled with density proportional to $|\langle x,y\rangle|$, and binary answers must have product $\operatorname{sgn}\langle x,y\rangle$. Its classical and quantum biases are $z_n$ and $1/(nz_n)$, respectively, where $z_n=\mathbb E|\langle x,y\rangle|\sim\sqrt{2/(\pi n)}$. For even $n=2m$, any finite-dimensional strategy achieving at least $(1-\varepsilon)$ times the optimal quantum bias admits one pair of local isometries certifying the entire family $\gamma(x)=\sum_i x_i\gamma_i$. Squared state error relative to $m$ maximally entangled qubit pairs plus an auxiliary state is $O(n^2\varepsilon)$; sphere-averaged squared, state-dependent measurement errors are $O(n\varepsilon)$. Both dimension dependences are optimal among uniform bounds linear in $\varepsilon$. Randomized products of the devices’ observables additionally certify the spin representation of $\mathrm{Pin}(n)$, with squared error $O(n^3\varepsilon)$ averaged over group elements and rotations. The proof derives approximate linearity, consistency and anticommutation from near-optimality, selects a random orthonormal frame, and applies a state-dependent Gowers–Hatami rounding argument to the finite Clifford group. Rotation averaging extends certification to products. Additional results establish classical and vector-strategy rigidity and give an explicit compact-group rounding construction. Certification is averaged, not pointwise: measurements on any null set can change without affecting the score. The main theorems assume finite-dimensional tensor-product strategies and even $n$; implementing the pin-group construction requires unitary access to the observables.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.