Study establishes theory of overparametrization in quantum neural networks A theoretical proof shows that a technique called overparametrization enhances performance in quantum machine learning for applications that stymie classical computers.