Distributional Deep Gaussian Processes

Sebastian Popescu
Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, PMLR 337:5504-5540, 2026.

Abstract

Deep {Gaussian} processes (DGPs) offer a principled {Bayesian} framework with hierarchical uncertainty propagation, but their reliable propagation of uncertainty and out-of-distribution ({OOD}) detection performance remains underexplored and often unreliable in safety-critical settings. In this work, we propose a novel kernel operating in both {Euclidean} and {Wasserstein}-2 space to better account for the geometry of representation learning spaces, thus circumventing a common pathology called feature collapse, whereby inliers and outliers get mapped to similar spaces. Empirically, our approach consistently improves {OOD} detection in convolutional image tasks and shows improved performance on tabular datasets.

Cite this Paper

BibTeX
@InProceedings{pmlr-v337-popescu26a, title = {Distributional Deep {Gaussian} Processes}, author = {Popescu, Sebastian}, booktitle = {Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence}, pages = {5504--5540}, year = {2026}, editor = {Perković, Emilija and Malinsky, Daniel}, volume = {337}, series = {Proceedings of Machine Learning Research}, month = {17--21 Aug}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v337/main/assets/popescu26a/popescu26a.pdf}, url = {https://proceedings.mlr.press/v337/popescu26a.html}, abstract = {Deep {Gaussian} processes (DGPs) offer a principled {Bayesian} framework with hierarchical uncertainty propagation, but their reliable propagation of uncertainty and out-of-distribution ({OOD}) detection performance remains underexplored and often unreliable in safety-critical settings. In this work, we propose a novel kernel operating in both {Euclidean} and {Wasserstein}-2 space to better account for the geometry of representation learning spaces, thus circumventing a common pathology called feature collapse, whereby inliers and outliers get mapped to similar spaces. Empirically, our approach consistently improves {OOD} detection in convolutional image tasks and shows improved performance on tabular datasets.} }
Endnote
%0 Conference Paper %T Distributional Deep Gaussian Processes %A Sebastian Popescu %B Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence %C Proceedings of Machine Learning Research %D 2026 %E Emilija Perković %E Daniel Malinsky %F pmlr-v337-popescu26a %I PMLR %P 5504--5540 %U https://proceedings.mlr.press/v337/popescu26a.html %V 337 %X Deep {Gaussian} processes (DGPs) offer a principled {Bayesian} framework with hierarchical uncertainty propagation, but their reliable propagation of uncertainty and out-of-distribution ({OOD}) detection performance remains underexplored and often unreliable in safety-critical settings. In this work, we propose a novel kernel operating in both {Euclidean} and {Wasserstein}-2 space to better account for the geometry of representation learning spaces, thus circumventing a common pathology called feature collapse, whereby inliers and outliers get mapped to similar spaces. Empirically, our approach consistently improves {OOD} detection in convolutional image tasks and shows improved performance on tabular datasets.
APA
Popescu, S.. (2026). Distributional Deep Gaussian Processes. Proceedings of the 42nd Conference on Uncertainty in Artificial Intelligence, in Proceedings of Machine Learning Research 337:5504-5540 Available from https://proceedings.mlr.press/v337/popescu26a.html.

Related Material