Experimental hybrid quantum-classical reinforcement learning by boson sampling: how to train a quantum cloner

Abstract

We report on experimental implementation of a machine-learned quantum gate driven by a classical control. The gate learns optimal phase-covariant cloning in a reinforcement learning scenario having fidelity of the clones as reward. In our experiment, the gate learns to achieve nearly optimal cloning fidelity allowed for this particular class of states. This makes it a proof of present-day feasibility and practical applicability of the hybrid machine learning approach combining quantum information processing with classical control. Moreover, our experiment can be directly generalized to larger interferometers where the computational cost of classical computer is much lower than the cost of boson sampling.

Variational Quantum Cloning: Improving Practicality for Quantum Cryptanalysis

“…We introduce three cost functions here suitable for approximate cloning tasks, one inspired by Ref. 66 , and the other two adapted from the literature on variational algorithms 30,37,44,46,50 . We begin by stating the functions, and then discussing the various ingredients and their relative advantages.…”

“…we have only two output parties, j ∈ {B, E}), and remove the expectation value over states, we recover the cost function used in Ref. 66 . A useful feature of this cost is that symmetry is explicitly enforced by the difference term (F i (θ) − F j (θ)) 2 .…”

“…In contrast, the work of Ref. 66 considers a black box, gradient-free training using the Nelder-Mead optimiser 86 . In this work, we opt for a gradient based approach and derive the analytic gradient for our cost functions.…”

“…We mention that the work of Ref. 66 considered a similar idea of using a variational quantum circuit to learn the parameters in optimal phase covariant cloning circuits. In this work, we significantly expand on this idea to propose optimal cloning circuits for a wide class of cloning problems including universal/state-dependent cloning under a symmetric/asymmetric framework.…”

Efficient Bipartite Entanglement Detection Scheme with a Quantum Adversarial Solver

“…In our proposal, the gate arrangements in U G and U D should be carefully designed, due to the tradeoff between the trainability and expressivity of the variational optimization algorithms [46][47][48]. To further improve the performance of our scheme, a promising research direction is applying variable quantum circuit strategies [49][50][51] to automatically seek an optimal gates arrangement and maximize the benefit while minimizing the number of circuit elements.…”