Adaptive Guidance Scaling for Posterior Diffusion-based Sampling

ICLR 2026 Conference Submission12640 Authors

Keywords: Diffusion models, Inverse Problems, Posterior Sampling

TL;DR: APS: posterior-guided diffusion sampler that sets the likelihood step size in an adaptive, robust manner; obtains state-of-the-art results without task-specific tuning.

Abstract:

Diffusion models have recently emerged as powerful generative priors for solving inverse problems, achieving state-of-the-art results across various imaging tasks. A central challenge in this setting lies in balancing the contribution of the prior with the data fidelity term: overly aggressive likelihood updates may introduce artifacts, while conservative updates can slow convergence or yield suboptimal reconstructions.
In this work, we propose an adaptive likelihood step-size strategy to guide the diffusion process for inverse-problem formulations. Specifically, we develop an observation-dependent weighting scheme based on the agreement between two different approximations of the intractable intermediate likelihood gradients, that adapts naturally to the diffusion schedule, time re-spacing, and injected stochasticity.
The resulting approach, Adaptive Posterior diffusion Sampling (APS), is hyperparameter-free and improves reconstruction quality across diverse imaging tasks—including super-resolution, Gaussian deblurring, and motion deblurring—on CelebA-HQ and ImageNet-256 validation sets. APS consistently surpasses existing diffusion-based baselines in perceptual quality without any task-specific tuning. Extensive ablation studies further demonstrate its robustness to the number of diffusion steps, observation noise levels, and varying stochasticity.

Primary Area: generative models

Submission Number: 12640