Wavefunction Flows: Efficient Quantum Simulation of Continuous Flow Models

2025-10-09

AI Summary

This work addresses the challenge of efficiently preparing quantum samples (qsamples)—coherent encodings of arbitrary probability distributions—on quantum computers. We establish a deep connection between classical continuous normalizing flows and quantum dynamics: the probabilistic transformation in flow models is mapped to unitary evolution governed by the Schrödinger equation, enabling explicit construction of the corresponding Hamiltonian. Leveraging this insight, we design a scalable quantum algorithm that integrates flow matching, diffusion modeling, and continuous-time Markov process principles to generate qsamples for broad distribution families—including non-Gaussian and multimodal distributions—with provable efficiency. This constitutes the first systematic framework translating classical generative models into quantum state preparation protocols. Our approach provides rigorous quantum advantage for statistical tasks such as mean estimation and property testing, overcoming longstanding limitations of oracle-dependent or distribution-restricted quantum sampling methods.

Technology Category

• Machine Learning: Quantum Machine Learning
• Search and Optimization: Sampling/Simulation-based Search
• Reasoning under Uncertainty: Probabilistic Programming

Application Category

• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Security and Privacy: Data transparency and provenance

Abstract

Flow models are a cornerstone of modern machine learning. They are generative models that progressively transform probability distributions according to learned dynamics. Specifically, they learn a continuous-time Markov process that efficiently maps samples from a simple source distribution into samples from a complex target distribution. We show that these models are naturally related to the Schr""odinger equation, for an unusual Hamiltonian on continuous variables. Moreover, we prove that the dynamics generated by this Hamiltonian can be efficiently simulated on a quantum computer. Together, these results give a quantum algorithm for preparing coherent encodings (a.k.a., qsamples) for a vast family of probability distributions--namely, those expressible by flow models--by reducing the task to an existing classical learning problem, plus Hamiltonian simulation. For statistical problems defined by flow models, such as mean estimation and property testing, this enables the use of quantum algorithms tailored to qsamples, which may offer advantages over classical algorithms based only on samples from a flow model. More broadly, these results reveal a close connection between state-of-the-art machine learning models, such as flow matching and diffusion models, and one of the main expected capabilities of quantum computers: simulating quantum dynamics.

Problem

• Efficient quantum simulation of continuous flow models
• Preparing quantum encodings for probability distributions
• Connecting machine learning flow models to quantum dynamics

Innovation

• Flow models map distributions via learned dynamics
• Quantum simulation of Hamiltonian enables efficient computation
• Reduces quantum state preparation to classical learning task