Generative Uncertainty in Diffusion Models

Metod Jazbec, Eliot Wong-Toi, Guoxuan Xia, Dan Zhang, Eric Nalisnick, Stephan Mandt

Keywords: diffusion models; uncertainty quantification; Bayesian inference

TL;DR: We propose generative uncertainty to detect low-quality samples in diffusion models.

Abstract: Diffusion models have recently driven significant breakthroughs in generative modeling. While state-of-the-art models produce high-quality samples on average, individual samples can still be low quality. Detecting such samples without human inspection remains a challenging task. To address this, we propose a Bayesian framework for estimating generative uncertainty of synthetic samples. We outline how to make Bayesian inference practical for large, modern generative models and introduce a new semantic likelihood (evaluated in the latent space of a feature extractor) to address the challenges posed by high-dimensional sample spaces. Through our experiments, we demonstrate that the proposed generative uncertainty effectively identifies poor-quality samples and significantly outperforms existing uncertainty-based methods. Notably, our Bayesian framework can be applied post-hoc to any pretrained diffusion or flow matching model (via the Laplace approximation), and we propose simple yet effective techniques to minimize its computational overhead during sampling.

Code Link: https://github.com/metodj/DIFF-UQ

Readers: auai.org/UAI/2025/Conference, auai.org/UAI/2025/Conference/Area_Chairs, auai.org/UAI/2025/Conference/Reviewers, auai.org/UAI/2025/Conference/Submission63/Authors, auai.org/UAI/2025/Conference/Submission63/Reproducibility_Reviewers

Submission Number: 63