Flexible Bayesian Last Layer Models Using Implicit Priors and Diffusion Posterior Sampling
This addresses the problem of limited expressive capacity in Bayesian neural networks for practitioners dealing with complex datasets, though it appears incremental as an enhancement to existing BLL frameworks.
The paper tackles the limitation of Bayesian Last Layer models' Gaussian priors on non-Gaussian or outlier-rich data by introducing a method combining diffusion techniques and implicit priors for variational learning, resulting in improved accuracy, calibration, and out-of-distribution detection.
Bayesian Last Layer (BLL) models focus solely on uncertainty in the output layer of neural networks, demonstrating comparable performance to more complex Bayesian models. However, the use of Gaussian priors for last layer weights in Bayesian Last Layer (BLL) models limits their expressive capacity when faced with non-Gaussian, outlier-rich, or high-dimensional datasets. To address this shortfall, we introduce a novel approach that combines diffusion techniques and implicit priors for variational learning of Bayesian last layer weights. This method leverages implicit distributions for modeling weight priors in BLL, coupled with diffusion samplers for approximating true posterior predictions, thereby establishing a comprehensive Bayesian prior and posterior estimation strategy. By delivering an explicit and computationally efficient variational lower bound, our method aims to augment the expressive abilities of BLL models, enhancing model accuracy, calibration, and out-of-distribution detection proficiency. Through detailed exploration and experimental validation, We showcase the method's potential for improving predictive accuracy and uncertainty quantification while ensuring computational efficiency.