Posterior sampling by source-space MCMC via prior-based few-step transport maps
Abstract
Bayesian inference increasingly uses informative but implicit priors represented only by samples, such as historical ensembles, simulator outputs, and pretrained generative models. The same computational problem appears in the test-time guidance task (generalized Bayes), where an explicit positive weight, e.g., an exponentiated reward, tilts an implicit prior. We develop a framework for source-space generalized Bayesian inference that combines inexpensive few-step prior transports with posterior stability guarantees. Specifically, we represent the prior using a one- or few-step improved MeanFlow (iMF) map and perform posterior sampling in its Gaussian source space. We establish Wasserstein error bounds between the exact and learned posteriors in terms of the joint population iMF and auxiliary-velocity loss, decomposed into training suboptimality and model-class approximation error. In the iMF source space, we adopt parallel tempering with preconditioned Crank-Nicolson updates and introduce a hybrid variant that incorporates split Hamiltonian Monte Carlo to improve sampling efficiency. Synthetic experiments show that the proposed framework can approximate posterior distributions accurately and efficiently, while CLIP-guided ImageNet experiments demonstrate its ability to steer a pretrained iMF image prior toward text-specified preferences.
Related
- Power-posterior implicit sampling for efficient Bayesian experimental design: Application to contaminant source identification in groundwater flow
- Flow Annealing Posterior Sampling for Function-Space Regression and Inverse Problems
- Bounds on the Posterior-to-Prior Ratios for Inclusion Belief under Bounded Differential Privacy
- On Prior-to-Posterior Stability in the Wasserstein Metric for Bayesian Inverse Problems
- Accelerating posterior sampling in Bayesian inverse problems with diffusion models: an online-offline perspective
BibTeX
@misc{luu2026posterior,
title = {{Posterior sampling by source-space MCMC via prior-based few-step transport maps}},
author = {Hoang Phuc Hau Luu and Marcelo Hartmann and Zhongjian Wang},
howpublished = {arXiv},
year = {2026},
month = oct,
doi = {10.48550/arxiv.2610.01034},
eprint = {2610.01034},
archivePrefix = {arXiv},
url = {https://arxiv.org/abs/2610.01034},
note = {CC BY 4.0},
}
RIS
TY - UNPB
TI - Posterior sampling by source-space MCMC via prior-based few-step transport maps
AU - Hoang Phuc Hau Luu
AU - Marcelo Hartmann
AU - Zhongjian Wang
PY - 2026
DA - 2026/10/01
PB - arXiv
DO - 10.48550/arxiv.2610.01034
UR - https://arxiv.org/abs/2610.01034
AB - Bayesian inference increasingly uses informative but implicit priors represented only by samples, such as historical ensembles, simulator outputs, and pretrained generative models. The same computational problem appears in the test-time guidance task (generalized Bayes), where an explicit positive weight, e.g., an exponentiated reward, tilts an implicit prior. We develop a framework for source-space generalized Bayesian inference that combines inexpensive few-step prior transports with posterior stability guarantees. Specifically, we represent the prior using a one- or few-step improved MeanFlow (iMF) map and perform posterior sampling in its Gaussian source space. We establish Wasserstein error bounds between the exact and learned posteriors in terms of the joint population iMF and auxiliary-velocity loss, decomposed into training suboptimality and model-class approximation error. In the iMF source space, we adopt parallel tempering with preconditioned Crank-Nicolson updates and introduce a hybrid variant that incorporates split Hamiltonian Monte Carlo to improve sampling efficiency. Synthetic experiments show that the proposed framework can approximate posterior distributions accurately and efficiently, while CLIP-guided ImageNet experiments demonstrate its ability to steer a pretrained iMF image prior toward text-specified preferences.
ER -