A groundbreaking examine has recognized a novel strategy to allow quantum computer systems to know and predict quantum programs with only some easy examples. The examine makes use of Quantum Neural Networks (QNNs), machine studying fashions that mimic quantum system conduct. In distinction to traditional studying fashions requiring quite a few examples, QNNs use a couple of ‘product states,’ that are less complicated and extra manageable types of quantum states.
Researchers have made a pivotal development in quantum computing by demonstrating how quantum neural networks can understand and predict quantum systems using a few simple ‘product states,’ potentially leading to more efficient and reliable quantum computers.
Imagine a world where computers can unravel the mysteries of quantum mechanics, enabling us to study the behavior of complex materials or simulate the intricate dynamics of molecules with unprecedented accuracy.
Thanks to a pioneering study led by Professor Zoe Holmes and her team at EPFL, we are now closer to that becoming a reality. Working with researchers at Caltech, the Free University of Berlin, and the Los Alamos National Laboratory, they have found a new way to teach a quantum computer how to understand and predict the behavior of quantum systems, even with a few simple examples.
Quantum neural networks (QNNs)
The researchers worked on “quantum neural networks” (QNNs), a type of machine-learning model designed to learn and process information using principles inspired by quantum mechanics in order to mimic the behavior of quantum systems.
Just like the neural networks used in artificial intelligence, QNNs are made of interconnected nodes, or “neurons,” that perform calculations. The difference is that, in QNNs, the neurons operate on the principles of quantum mechanics, allowing them to handle and manipulate quantum information.
“Normally, when we teach a computer something, we need a lot of examples,” says Holmes. “But in this study, we show that with just a few simple examples called ‘product states’ the computer can learn how a quantum system behaves even when dealing with entangled states, which are more complicated and challenging to understand.”
Product states
The ‘product states’ that the scientists used refer to a concept in quantum mechanics that describes the specific type of state for a quantum system. For example, if a quantum system is composed of two electrons, then its product state is formed when each individual electron’s state is considered independently and then combined.
Product states are often used as a starting point in quantum computations and measurements because they provide a simpler and more manageable framework for studying and understanding the behavior of quantum systems before moving on to more complex and entangled states where the particles are correlated and cannot be described independently.
Better quantum computers ahead
The researchers demonstrated that by training QNNs using only a few of these simple examples, computers can effectively grasp the complex dynamics of entangled quantum systems.
Holmes explains: “This means that might be able to learn about and understand quantum systems using smaller, simpler computers, like the near-term intermediary scale [NISQ] computer systems we’re prone to have within the coming years, as a substitute of needing giant and complicated ones, which can be a long time away.”
The work additionally opens up new potentialities for utilizing quantum computer systems to resolve essential issues like learning complicated new supplies or simulating the conduct of molecules.
Lastly, the tactic improves the efficiency of quantum computer systems by enabling the creation of shorter and extra error-resistant applications. By studying how quantum programs behave, we will streamline the programming of quantum computer systems, resulting in improved effectivity and reliability. “We will make quantum computer systems even higher by making their applications shorter and fewer liable to errors,” says Holmes.
Reference: “Out-of-distribution generalization for studying quantum dynamics” by Matthias C. Caro, Hsin-Yuan Huang, Nicholas Ezzell, Joe Gibbs, Andrew T. Sornborger, Lukasz Cincio, Patrick J. Coles and Zoë Holmes, 5 July 2023, Nature Communications.
DOI: 10.1038/s41467-023-39381-w
Funding: Technical University of Munich, Elite Network of Bavaria, Studienstiftung des Deutschen Volkes, BMWi, U.S. Department of Energy, Google, Los Alamos National Laboratory, Sandoz Family Foundation-Monique de Meuron program