Quantum latent distributions in deep generative models

Many successful families of generative models leverage a low-dimensional latent distribution that is mapped to a data distribution. Though simple latent distributions are often used, the choice of distribution has a strong impact on model performance. Recent experiments have suggested that the probability distributions produced by quantum processors, which are typically highly correlated and classically intractable, can lead to improved performance on some datasets. However, when and why latent distributions produced by quantum processors can improve performance, and whether these improvements are connected to quantum properties of these distributions, are open questions that we investigate in this work. We show in theory that, under certain conditions, these "quantum latent distributions" enable generative models to produce data distributions that classical latent distributions cannot efficiently produce. We provide intuition as to the underlying mechanisms that could explain a performance advantage on real datasets. Based on this, we perform extensive benchmarking on a synthetic quantum dataset and the QM9 molecular dataset, using both simulated and real photonic quantum processors. We find that the statistics arising from quantum interference lead to improved generative performance compared to classical baselines, suggesting that quantum processors can play a role in expanding the capabilities of deep generative models.

AI generators work by transforming simple input data into realistic outputs. However, just as a sculptor is limited by their clay, these models are constrained by the common use of excessively simple inputs. For complex target data rooted in physical or chemical processes, simple input data is not enough. We investigated whether quantum computers could serve as a better starting point. Quantum computers generate data through quantum interference, which is statistically richer than anything a standard computer can easily replicate. We showed mathematically that using data from quantum computers as inputs to AI generators can help them generate target data that classical computers cannot reach. We then tested this on two datasets: one drawn from a quantum process, and one containing drug-like molecules governed by quantum chemistry. We found that both simulated and real quantum hardware consistently outperformed classical alternatives. These results suggest quantum computers could become a practical tool for improving AI-generated content in science and beyond.