Many-body variational state preparation in the age of machine learning and quantum computing

by Prof. Giuseppe Carleo (EPFL)

Friday 29 Oct 2021, 11:00 → 12:00 Europe/Zurich

WHGA/001

In this seminar I will discuss several approaches to the fundamental problem of efficiently preparing many-body quantum states. I will present both classical and quantum algorithms for this task, focusing on the respective advantages and limitations.

In the context of quantum algorithms, I will consider variational states based on parameterised quantum circuits. 

I will introduce the concept of Quantum Natural Gradient [1] and its efficient implementation [2] using the Simultaneous Perturbation Stochastic Approximation. 

I will also discuss an efficient variational quantum algorithm named “projected – Variational Quantum Dynamics” (p-VQD) realizing an iterative, global projection of the exact time evolution onto the parameterized manifold [3]. 

In the context of classical algorithms, I will show instead how variational parameterisations based on neural network quantum states [4] can be used to simulate NISQ-scale quantum computers, and show the example of QAOA [5]. 

[1] James Stokes, Josh Izaac, Nathan Killoran, and Giuseppe Carleo, Quantum 4, 269 (2020)
[2] Julien Gacon, Christa Zoufal, Giuseppe Carleo, and Stefan Woerner, arXiv:2103.09232 (2021) 
[3] Stefano Barison, Filippo Vicentini, and Giuseppe Carleo, arXiv:2101.04579 (2021)
[4] Giuseppe Carleo, and Matthias Troyer, Science 355, 602 (2017)
[5] Matija Medvidović, and Giuseppe Carleo, npj Quantum Inf 7, 101 (2021)

SCD_coll_20211029.pdf