Quantum circuit architecture search on a superconducting processor

TL;DR

This paper demonstrates the first use of quantum architecture search to automatically design efficient ansatze for variational quantum algorithms on a superconducting processor, significantly improving classification accuracy.

Contribution

It introduces an automatic ansatz design technique, quantum architecture search, tailored for hardware-efficient quantum classifiers on NISQ devices, showing substantial accuracy improvements.

Findings

QAS improves test accuracy from 31% to 98%.

Visualization of loss landscape explains performance gains.

Provides guidance for large-scale quantum learning.

Abstract

Variational quantum algorithms (VQAs) have shown strong evidences to gain provable computational advantages for diverse fields such as finance, machine learning, and chemistry. However, the heuristic ansatz exploited in modern VQAs is incapable of balancing the tradeoff between expressivity and trainability, which may lead to the degraded performance when executed on the noisy intermediate-scale quantum (NISQ) machines. To address this issue, here we demonstrate the first proof-of-principle experiment of applying an efficient automatic ansatz design technique, i.e., quantum architecture search (QAS), to enhance VQAs on an 8-qubit superconducting quantum processor. In particular, we apply QAS to tailor the hardware-efficient ansatz towards classification tasks. Compared with the heuristic ansatze, the ansatz designed by QAS improves test accuracy from 31% to 98%. We further explain this superior performance by visualizing the loss landscape and analyzing effective parameters of all ansatze. Our work provides concrete guidance for developing variable ansatze to tackle various large-scale quantum learning problems with advantages.