QLunch: Emilio Onorati

Speaker: Emilio Onorati from TUM

Title: Random quantum circuits and classical shadows for noise characterization

Abstract; In recent years we are observing important progress towards the realization of large scale quantum computers, yet noise in quantum hardware is still one of the main obstacles towards fully functional quantum computation. Strategies to accurately characterize the error affecting experimentally implemented quantum processes can help in calibrating quantum architectures and in tailoring error correction and mitigation techniques to the actual noise in the device – hence drastically reducing the resource cost of these procedures.

 In this talk we will present a family (referred to as randomized benchmarking) of error characterization procedures based on random quantum circuits. These methods are among the few ones that are suited to investigate noise in large-scale quantum computers in a scalable fashion, and possess further desired properties at the cost of additional assumptions. We will first discuss how some of its limitations can be lifted by leveraging symmetry properties of quantum gates https://arxiv.org/abs/1811.11775 . Moreover, we will show how the numerous family of randomized benchmarking protocols can be unified under a common mathematical description using the formalism of Fourier transforms on irreducible representations https://arxiv.org/abs/2010.07974 .

 In the second part of the talk we will consider a recent procedure, classical shadow estimation, which allows us to retrieve many properties of an unknown quantum state with few measurements. Inspired by this technique, we will present a method to efficiently characterize the structure of noise influencing experimental quantum gate sets https://arxiv.org/abs/2110.13178 .