Arbitrary Polynomial Separations in Trainable Quantum Machine Learning

========

Abstract

=========

Recent theoretical results in quantum machine learning have demonstrated a general trade-off between the expressive power of quantum neural networks (QNNs) and their trainability; as a corollary of these results, practical exponential separations in expressive power over classical machine learning models are believed to be infeasible, as such QNNs generically take a time to train that is exponential in the model size. We here circumvent these negative results by constructing a hierarchy of efficiently trainable QNNs that exhibit unconditionally provable, polynomial memory separations of arbitrary constant degree over classical neural networks—including state-of-the-art models, such as Transformers—in performing a classical sequence modeling task. This construction is also computationally efficient, as each unit cell of the introduced class of QNNs only has constant gate complexity. We show that contextuality—informally, a quantitative notion of semantic ambiguity—is the source of the expressivity separation, suggesting that other learning tasks with this property may be a natural setting for the use of quantum learning algorithms.

 

=================

About the Speaker

=================

Eric is a Burke Fellow at Caltech who recently completed his PhD at MIT under the joint supervision of Aram Harrow and Misha Lukin. Much of his research involves studying the limitations of quantum algorithms through the lens of statistical physics and quantum foundations theory. These insights have led to the development of novel quantum algorithms for optimization and learning, and a new, physically motivated approach to quantum computational complexity.