Given many copies of an unknown quantum state
ρ, we consider the task of learning a classical description of its principal eigenstate. Namely, assuming that
ρ has an eigenstate
∣ϕ⟩ with (unknown) eigenvalue
λ>1/2, the goal is to learn a (classical shadows style) classical description of
∣ϕ⟩ which can later be used to estimate expectation values
⟨ϕ∣O∣ϕ⟩ for any
O in some class of observables. We consider the sample-complexity setting in which generating a copy of
ρ is expensive, but joint measurements on many copies of the state are possible. We present a protocol for this task scaling with the principal eigenvalue
λ and show that it is optimal within a space of natural approaches, e.g., applying quantum state purification followed by a single-copy classical shadows scheme. Furthermore, when
λ is sufficiently close to
1, the performance of our algorithm is optimal--matching the sample complexity for pure state classical shadows.
Quantum computers and simulators promise to enable the study of strongly correlated quantum systems. Yet, surprisingly, it is hard for them to compute ground states. They can, however, efficiently compute the dynamics of closed quantum systems. We propose a method to study the quantum thermodynamics of strongly correlated electrons from quantum dynamics. We define time-averaged classical shadows (TACS) and prove it is a classical shadow(CS) of the von Neumann ensemble, the time-averaged density matrix. We then show that the diffusion maps, an unsupervised machine learning algorithm, can efficiently learn the phase diagram and phase transition of the one-dimensional transverse field Ising model both for ground states using CS and state trajectories using TACS. It does so from state trajectories by learning features that appear to be susceptibility and entropy from a total of 90 000 shots taken along a path in the microcanonical phase diagram. Our results suggest a low number of shots from quantum simulators can produce quantum thermodynamic data with a quantum advantage.
Classical shadows constitute a protocol to estimate the expectation values of a collection of M observables acting on O(1) qubits of an unknown n-qubit state with a number of measurements that is independent of n and that grows only logarithmically with M. We propose a local variant of the quantum Wasserstein distance of order 1 of De Palma et al. [IEEE Trans. Inf. Theory 67, 6627–6643 (2021)] and prove that the classical shadow obtained measuring O(log n) copies of the state to be learned constitutes an accurate estimate with respect to the proposed distance. We apply the results to quantum generative adversarial networks, showing that quantum access to the state to be learned can be useful only when some prior information on such state is available.