部分转置下纯度测试与乘积测试的紧界
Tight Bounds for Purity and Product Testing from Partial Transposition
浏览论文内容
中文总结 AI 辅助
本文针对纯度测试与乘积测试任务，证明正部分转置（PPT）松弛的渐近样本复杂度下界与非自适应单副本协议匹配，为自适应单副本测量提供了基于对称子空间恒等式的紧下界推导方法。
中文摘要 AI 辅助
对未知量子态多副本的相干测量可大幅减少学习其性质所需的样本数，但实验上仍具挑战性。当前实验通常每次制备并测量一个副本，可能根据早期结果调整后续测量，核心挑战是适应性，这使得可能的测量策略空间难以表征。正部分转置（PPT）松弛通过考虑更大、数学上易处理的测量类别绕过该复杂性，代价是可能弱化所得界。本文出人意料地表明，对于纯度测试和乘积测试这两个基础任务，该松弛在渐近样本复杂度层面未损失任何性能：针对完整PPT测量类别的下界与非自适应单副本协议匹配。此外，我们的证明仅需基本对称子空间恒等式，为自适应单副本测量的紧下界提供了简单途径。
英文摘要
Coherent measurements across multiple copies of an unknown quantum state can substantially reduce the number of samples required to learn its properties, but remain experimentally challenging. Current experiments typically prepare and measure one copy at a time, potentially adapting later measurements to earlier outcomes. A central challenge is adaptivity, which makes the space of possible measurement strategies difficult to characterize. The positive-partial-transpose (PPT) relaxation bypasses this complexity by considering a larger, mathematically tractable class of measurements, at the risk of weakening the resulting bounds. Here we show, surprisingly, that the relaxation loses nothing at the level of asymptotic sample complexity for two fundamental tasks: purity testing and product testing. In both cases, lower bounds against the full class of PPT measurements are matched by nonadaptive single-copy protocols. Moreover, our proof requires only basic symmetric subspace identities, providing a simple route to sharp lower bounds for adaptive single-copy measurements.
