What can be selected after collecting a classical shadow?

Question: Which capability distinguishes a reusable classical shadow from a measurement protocol tailored to one fixed observable? A) It reconstructs the exact density matrix with fewer than one measurement per matrix element B) It allows unlimited observable estimation with constant error regardless of observable complexity C) It stores a coherent quantum copy that can later be measured in arbitrary bases D) It permits target observables to be chosen after the randomized measurements have already been completed Correct: D Explanation: A classical shadow is classical measurement data together with a reconstruction rule. The same dataset can be reused to estimate many properties, including properties that were not specified before data collection, provided the shadow contains enough information for those observables. Topic: advanced quantum computing / classical shadows / quantum learning