arXiv:2610.06794
Testing bosonic and fermionic Gaussian states with a constant number of copies
Jop Briët, Davi Castro-Silva, Michael Walter, Freek Witteveen
Abstract
Bosonic and fermionic Gaussian quantum states are families of quantum states that play a central role in quantum physics and quantum information processing. We study the problem of testing whether a given state is either Gaussian, or has bounded overlap with any Gaussian state. Previous analyses of this testing problem led to tests that require a number of copies of the state that scales with the number of modes (the system size). In this work we show that a constant number of copies of the state suffices. We derive these results based on the symmetries of identical copies of Gaussian states, and by bounding the distance to the set of Gaussian states through a gradient flow and a spectral bound studied previously for the Kac random walk on a sphere.
AI Summary
The paper asks whether an unknown pure state can be tested for bosonic or fermionic Gaussianity with a copy complexity independent of the number of modes $m$. It gives perfectly complete tests distinguishing Gaussian states from states with squared overlap at most $1-\varepsilon$ with every pure Gaussian state, using $O(\varepsilon^{-1}\log(1/\delta))$ copies and failure probability at most $\delta$. Previous bounds discussed in the paper scale with $m$. The fermionic and centered-bosonic tests act on two copies, projecting onto the kernel of a generator of rotations between copies. A two-copy coherent-state test interferes identical inputs on a balanced beam splitter and checks one output for vacuum. General bosonic Gaussianity is tested on four copies by combining beam-splitter outputs with the centered test. If a single test accepts with probability at least $1-\eta$, the squared fidelity with the target class is at least $1-8\eta$ for fermionic or centered-bosonic Gaussian states, $1-4\eta$ for coherent states, and $1-48\eta$ for general bosonic Gaussian states. The robustness analysis introduces a general gradient-flow argument connecting a two-copy accepting projector $P$ to the spectrum of $S=(P_{12}+P_{13}+P_{23})/3$. For Gaussianity, the relevant spectrum lies in $[0,7/12]\cup\{1\}$, obtained from the $5/12$ spectral gap of the Kac walk through an $\mathrm{SO}(3)$ representation. The testing guarantees concern pure inputs and exact membership versus a fidelity-separated alternative, rather than testing general mixed Gaussian states. Constant copy complexity does not imply mode-independent implementation cost. Closely related independent concurrent work is noted.
The authors thought about the problem for quite some time, and eventually it was solved with important contributions by an LLM.
Central claim may not hold: The abstract does not specify that the unknown input must be pure, whereas the paper's testing guarantees cover pure inputs and do not establish Gaussianity testing for arbitrary mixed states.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 19 pages, 1 figure