X-EviProbe: Post-hoc Parameter-Free Evidential Uncertainty Quantification for Frozen Graph Neural Networks

Chenghua Guo, Sihong Xie, Xi Zhang
Proceedings of the 43rd International Conference on Machine Learning, PMLR 306:38534-38561, 2026.

Abstract

Reliable uncertainty quantification (UQ) is crucial for deploying graph neural networks (GNNs) in safety-critical settings, yet dominant solutions either rely on costly multi-pass sampling or require retraining—often using black-box auxiliary models—to obtain evidential semantics. We propose X-EviProbe, a simple and parameter-free post-hoc framework that turns a frozen GNN into an evidential predictor with a decomposable view of epistemic vs. aleatoric uncertainty. X-EviProbe constructs class-wise Dirichlet evidence by probing the frozen latent space and the model’s native outputs, and incorporates graph structure via lightweight evidence-strength propagation. This yields a transparent evidential representation without retraining or additional neural components. Extensive experiments on seven benchmarks show that X-EviProbe consistently ranks among the top methods for both OOD detection and misclassification detection, improving AUROC by up to 33.4% and 8.7% over the strongest baselines.

Cite this Paper

BibTeX
@InProceedings{pmlr-v306-guo26ac, title = {X-{E}vi{P}robe: Post-hoc Parameter-Free Evidential Uncertainty Quantification for Frozen Graph Neural Networks}, author = {Guo, Chenghua and Xie, Sihong and Zhang, Xi}, booktitle = {Proceedings of the 43rd International Conference on Machine Learning}, pages = {38534--38561}, year = {2026}, editor = {Zhang, Tong and Dudik, Miroslav and Jaggi, Martin and Agarwal, Alekh and Li, Sharon and Schuurmans, Dale and Zhu, Jerry and Berkenkamp, Felix and Dong, Hanze and Bietti, Alberto}, volume = {306}, series = {Proceedings of Machine Learning Research}, month = {06--11 Jul}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v306/main/assets/guo26ac/guo26ac.pdf}, url = {https://proceedings.mlr.press/v306/guo26ac.html}, abstract = {Reliable uncertainty quantification (UQ) is crucial for deploying graph neural networks (GNNs) in safety-critical settings, yet dominant solutions either rely on costly multi-pass sampling or require retraining—often using black-box auxiliary models—to obtain evidential semantics. We propose X-EviProbe, a simple and parameter-free post-hoc framework that turns a frozen GNN into an evidential predictor with a decomposable view of epistemic vs. aleatoric uncertainty. X-EviProbe constructs class-wise Dirichlet evidence by probing the frozen latent space and the model’s native outputs, and incorporates graph structure via lightweight evidence-strength propagation. This yields a transparent evidential representation without retraining or additional neural components. Extensive experiments on seven benchmarks show that X-EviProbe consistently ranks among the top methods for both OOD detection and misclassification detection, improving AUROC by up to 33.4% and 8.7% over the strongest baselines.} }
Endnote
%0 Conference Paper %T X-EviProbe: Post-hoc Parameter-Free Evidential Uncertainty Quantification for Frozen Graph Neural Networks %A Chenghua Guo %A Sihong Xie %A Xi Zhang %B Proceedings of the 43rd International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2026 %E Tong Zhang %E Miroslav Dudik %E Martin Jaggi %E Alekh Agarwal %E Sharon Li %E Dale Schuurmans %E Jerry Zhu %E Felix Berkenkamp %E Hanze Dong %E Alberto Bietti %F pmlr-v306-guo26ac %I PMLR %P 38534--38561 %U https://proceedings.mlr.press/v306/guo26ac.html %V 306 %X Reliable uncertainty quantification (UQ) is crucial for deploying graph neural networks (GNNs) in safety-critical settings, yet dominant solutions either rely on costly multi-pass sampling or require retraining—often using black-box auxiliary models—to obtain evidential semantics. We propose X-EviProbe, a simple and parameter-free post-hoc framework that turns a frozen GNN into an evidential predictor with a decomposable view of epistemic vs. aleatoric uncertainty. X-EviProbe constructs class-wise Dirichlet evidence by probing the frozen latent space and the model’s native outputs, and incorporates graph structure via lightweight evidence-strength propagation. This yields a transparent evidential representation without retraining or additional neural components. Extensive experiments on seven benchmarks show that X-EviProbe consistently ranks among the top methods for both OOD detection and misclassification detection, improving AUROC by up to 33.4% and 8.7% over the strongest baselines.
APA
Guo, C., Xie, S. & Zhang, X.. (2026). X-EviProbe: Post-hoc Parameter-Free Evidential Uncertainty Quantification for Frozen Graph Neural Networks. Proceedings of the 43rd International Conference on Machine Learning, in Proceedings of Machine Learning Research 306:38534-38561 Available from https://proceedings.mlr.press/v306/guo26ac.html.

Related Material