Online Posterior Sampling with a Diffusion Prior
Branislav Kveton, Boris N. Oreshkin, Youngsuk Park, Aniket Anand Deshmukh, Rui Song
Abstract
Posterior sampling in contextual bandits with a Gaussian prior can be implemented exactly or approximately using the Laplace approximation. The Gaussian prior is computationally efficient but it cannot describe complex distributions. In this work, we propose approximate posterior sampling algorithms for contextual bandits with a diffusion model prior. The key idea is to sample from a chain of approximate conditional posteriors, one for each stage of the reverse diffusion process, which are obtained by the Laplace approximation. Our approximations are motivated by posterior sampling with a Gaussian prior, and inherit its simplicity and efficiency. They are asymptotically consistent and perform well empirically on a variety of contextual bandit problems.
BibTeX
@inproceedings{
kveton2024online,
title={Online Posterior Sampling with a Diffusion Prior},
author={Branislav Kveton and Boris N. Oreshkin and Youngsuk Park and Aniket Anand Deshmukh and Rui Song},
booktitle={The Thirty-eighth Annual Conference on Neural Information Processing Systems},
year={2024},
url={https://openreview.net/forum?id=7v0UyO0B6q}
}