Amortized Bayesian Workflow (Extended Abstract)

Marvin Schmitt, Chengkun LI, Aki Vehtari, Luigi Acerbi, Paul-Christian Bürkner, Stefan T. Radev

Keywords: Bayesian workflow, amortized inference

TL;DR: An adaptive Bayesian workflow that combines fast amortized inference with accurate MCMC, using diagnostics to optimize speed-accuracy trade-offs across thousands of datasets.

Abstract: Bayesian inference often faces a trade-off between computational speed and sampling accuracy. We propose an adaptive workflow that integrates rapid amortized inference with gold-standard MCMC techniques to achieve both speed and accuracy when performing inference on many observed datasets. Our approach uses principled diagnostics to guide the choice of inference method for each dataset, moving along the Pareto front from fast amortized sampling to slower but guaranteed-accurate MCMC when necessary. By reusing computations across steps, our workflow creates synergies between amortized and MCMC-based inference. We demonstrate the effectiveness of this integrated approach on a generalized extreme value task with 1000 observed data sets, showing efficiency gains (90x faster inference) while maintaining high posterior quality.

Submission Number: 9