Preconditioned One-Step Generative Modeling for Bayesian Inverse Problems in Function Spaces
This addresses computational bottlenecks in Bayesian inference for PDE-based problems, offering a fast alternative to MCMC, though it is incremental in improving generative modeling stability.
The paper tackles the instability of generative models in Bayesian inverse problems for PDEs by using a prior-aligned anisotropic Gaussian reference distribution, achieving stable inference and generating posterior samples in ~0.001 seconds.
We propose a machine-learning algorithm for Bayesian inverse problems in the function-space regime based on one-step generative transport. Building on the Mean Flows, we learn a fully conditional amortized sampler with a neural-operator backbone that maps a reference Gaussian noise to approximate posterior samples. We show that while white-noise references may be admissible at fixed discretization, they become incompatible with the function-space limit, leading to instability in inference for Bayesian problems arising from PDEs. To address this issue, we adopt a prior-aligned anisotropic Gaussian reference distribution and establish the Lipschitz regularity of the resulting transport. Our method is not distilled from MCMC: training relies only on prior samples and simulated partial and noisy observations. Once trained, it generates a $64\times64$ posterior sample in $\sim 10^{-3}$s, avoiding the repeated PDE solves of MCMC while matching key posterior summaries.