The method, developed at the University of Innsbruck, generates quantum circuits based on a user's specifications and tailors them to the capabilities of the quantum hardware on which they will run. Credit: University of Innsbruck/Harald Ritsch
The researchers Machine Learning A model for generating quantum circuits from text descriptions. It is similar to image generation methods such as the stable diffusion model. This method allows quantum computing.
One of the most important recent developments in machine learning (ML) is generative models, such as diffusion models. This includes: Stable diffusion and Dalare revolutionizing the field of image generation: these models are able to generate high-quality images based on text descriptions.
“Our new model for programming quantum computers does the same thing, but instead of generating images, it generates quantum circuits based on textual descriptions of the quantum operations to be performed,” the university's theory says. Gorka Muñoz Gil from the Department of Physics explains. Born in Innsbruck, Austria.
The challenges of quantum computing
To prepare a particular quantum state or run an algorithm on a quantum computer, we need to find an appropriate sequence of quantum gates to perform such an operation. While this is relatively easy in classical computing, it becomes a major challenge in quantum computing due to the peculiarities of the quantum world. Recently, many scientists have proposed ways to build quantum circuits using a number of dependent machine learning techniques. However, training these ML models is often very difficult because the machine has to simulate quantum circuits when learning. Diffusion models avoid such problems due to the way they are trained.
“This offers enormous advantages,” explains Gorka Muñoz Gil, who developed the new method together with Hans J. Briegel and Florian Fürter. “Moreover, we have shown that the denoising diffusion model is both accurate in its generation and very flexible, allowing the generation of circuits with different numbers of qubits, and types and numbers of quantum gates.”
The model can also be customized to prepare circuits that take into account quantum hardware connectivity, or how qubits are connected within a quantum computer.
“Once the model is trained, it is very cheap to create new circuits, which can then be used to discover new insights into the quantum operations of interest,” says Gorka Muñoz-Gil, citing another possible outcome of the new method.
quantum circuit generation
Developed at the University of Innsbruck, the technique generates quantum circuits based on a user's specifications and tailors them to the capabilities of the quantum hardware on which they will run. This is a major step towards fully unlocking the limits of quantum computing. The research is currently being carried out nature machine intelligence Additionally, the project received financial support from the Austrian Science Fund FWF and the European Union.
References: Florian Fürrutter, Gorka Muñoz-Gil, and Hans J. Briegel, “Quantum Circuit Synthesis with Diffusion Models,” May 20, 2024. Nature Machine Intelligence.
Publication date: 10.1038/s42256-024-00831-9
