arXiv:2609.39113
Optimal testing of fermionic and bosonic Gaussian states
Vishnu Iyer, Bryan O'Gorman, Ojas Parekh, Kevin Thompson, Andrew Zhao
Abstract
We study the problem of property testing quantum states: given a pure state $|ψ\rangle$ that either (A) belongs to some class $C$ of pure states, or (B) is $\varepsilon$-far in trace distance from all states in $C$, determine which is the case with high probability. It is known that there exists classes $C$ for which the sample complexity of property testing necessarily scales with the size of the system. In this work, we identify classes for which testing only requires a size-independent number of samples. We prove a tight sample complexity of $Θ(1/\varepsilon^2)$ for the property testing of five important classes of Gaussian quantum states: 1) Fermionic Gaussian states, 2) Slater determinants, 3) bosonic Gaussian states, 4) zero-mean bosonic Gaussian states, and 5) bosonic coherent states. While property testers for some of these classes have been studied previously, those analyses only gave upper bounds scaling polynomially in the number of modes. In each case, our tester is the projection onto the ``top'' irrep of two or three copies. We get to constant sample complexity by lower bounding how quickly the rejection probability increases with distance from the class; the main technical component involves bounding the spectral gap of symmetric extension of the test projector. We then establish optimality by proving matching lower bounds based on binary state discrimination for each class. Our testers and optimality claims also extend to the tolerant setting, wherein $ψ$ may be $O(\varepsilon)$-close to $C$ in case (A), rather than lying exactly inside.
AI Summary
The paper asks whether Gaussian structure in an unknown pure quantum state can be tested with a number of copies independent of the number of modes. It considers five classes: fermionic Gaussian states, Slater determinants, bosonic Gaussian states, zero-mean bosonic Gaussian states, and coherent states. For each class, it establishes optimal sample complexity $\Theta(1/\varepsilon^2)$ at constant success probability for distinguishing membership from trace distance at least $\varepsilon$ from the class. The protocols repeatedly measure two copies jointly, except for general bosonic Gaussian states, which require three. The tests project onto representation-theoretic “top” sectors; their perfect-completeness characterizations were largely known, while the analysis establishes dimension-independent soundness. The central framework relates rejection probability $q(\psi)$ to distance through $\alpha\,\mathrm{dist}(\psi,\mathcal C)^2\leq q(\psi)\leq k\,\mathrm{dist}(\psi,\mathcal C)^2$. It obtains this inequality from spectral gaps of symmetrically extended test operators, using a gradient-flow argument on projective Hilbert space. Representation theory and explicit diagonalization supply the required gaps. Matching lower bounds follow from binary discrimination between constructed nearby states. Tolerant testing uses $O(\log(1/\delta)/(\varepsilon_2-\beta\varepsilon_1)^2)$ copies, provided $\varepsilon_2>\beta\varepsilon_1$, with class-dependent constant $\beta$; thus the tolerant bounds do not cover arbitrary promise gaps. The results assume pure inputs, not general mixed Gaussian-state testing, and constant copy complexity does not mean constant total computation. The paper also notes simultaneous independent work establishing dimension-independent testing for the same classes.
In response, GPT 5.6 proposed a proof technique resembling the spectral gap technique outlined in \cref{sec:framework}, along with a candidate proof for the spectral gap for the fermionic Gaussian tester.
Error found: The introductory lower-bound theorem omits the ε₂ ≤ 1/√2 range established by its proof and nontrivial-size qualifications, and cannot apply to fermionic systems where every physical pure state is Gaussian. The gradient-inequality corollary incorrectly writes −7r for k = 3 and m = 7, whereas the proved formula gives −9r. The explicit fermionic eigenvalue and diagonalization formulas contain an extra factor of two relative to the defined Λ, without changing the tester's kernel. The alternative fermionic spectral-gap proof derives values inconsistent with its claimed spectrum and ends without establishing its proposed auxiliary identities.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.
- Comments
- 91 pages, 1 table