Approximate supervised learning of quantum gates via ancillary qubits

Abstract

We present strategies for the training of a qubit network aimed at the ancilla-assisted synthesis of multi-qubit gates based on a set of restricted resources. By assuming the availability of only time-independent single and two-qubit interactions, we introduce and describe a supervised learning strategy implemented through momentum-stochastic gradient descent with automatic differentiation methods. We demonstrate the effectiveness of the scheme by discussing the implementation of nontrivial three qubit operations, including a QFT and a half-adder gate.

Financial Policy for Promoting Equality Education in Rural Areas: Two Key Policy Implementation in China

“…Another topic is the noise effect. There are inspiring discussions on the noisy training data [77], robustness in tomographic completeness [78] and benefits of quantum noise [79]. In our scheme the influence of time independent system noise on T cancels out by a direct observation of Eq.…”

Quantum arithmetics via computation with minimized external control: The half-adder

“…Nonetheless, since the matrix logarithm is a multivalue function, for degenerate gates one can drastically simplify the resulting Hamiltonian and remove higher order interactions [38]. For nondegenerate gates, these matrix logarithms normally give rise to three-body interaction terms (or even of higher order), which are undesirable, as they unlikely appear naturally in physical systems and must be engineered via a superexchange mechanism [39].…”