Structural Alignment Improves Graph Test-Time Adaptation

Hans Hao-Hsun Hsu, Shikun Liu, Han Zhao, Pan Li
Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, PMLR 300:829-837, 2026.

Abstract

Graph-based learning excels at capturing interaction patterns in diverse domains like recommendation, fraud detection, and particle physics. However, its performance often degrades under distribution shifts, especially those altering network connectivity. Current methods to address these shifts typically require retraining with the source dataset, which is often infeasible due to computational or privacy limitations. We introduce Test-Time Structural Alignment (TSA), a novel algorithm for Graph Test-Time Adaptation (GTTA) that adapts a pretrained model to align graph structures during inference without the cost of retraining. Grounded in a theoretical understanding of graph data distribution shifts, TSA employs three synergistic strategies: uncertainty-aware neighborhood weighting to accommodate neighbor label distribution shifts, adaptive balancing of self-node and aggregated neighborhood representations based on their signal-to-noise ratio, and decision boundary refinement to correct residual label and feature shifts. Extensive experiments on synthetic and real-world datasets demonstrate TSA’s consistent outperformance of both non-graph TTA methods and state-of-the-art GTTA baselines.

Cite this Paper

BibTeX
@InProceedings{pmlr-v300-hsu26a, title = { Structural Alignment Improves Graph Test-Time Adaptation }, author = {Hsu, Hans Hao-Hsun and Liu, Shikun and Zhao, Han and Li, Pan}, booktitle = {Proceedings of The 29th International Conference on Artificial Intelligence and Statistics}, pages = {829--837}, year = {2026}, editor = {Khan, Emtiyaz and Li, Yingzhen and Solin, Arno and Ramdas, Aaditya}, volume = {300}, series = {Proceedings of Machine Learning Research}, month = {02--05 May}, publisher = {PMLR}, pdf = {https://raw.githubusercontent.com/mlresearch/v300/main/assets/hsu26a/hsu26a.pdf}, url = {https://proceedings.mlr.press/v300/hsu26a.html}, abstract = { Graph-based learning excels at capturing interaction patterns in diverse domains like recommendation, fraud detection, and particle physics. However, its performance often degrades under distribution shifts, especially those altering network connectivity. Current methods to address these shifts typically require retraining with the source dataset, which is often infeasible due to computational or privacy limitations. We introduce Test-Time Structural Alignment (TSA), a novel algorithm for Graph Test-Time Adaptation (GTTA) that adapts a pretrained model to align graph structures during inference without the cost of retraining. Grounded in a theoretical understanding of graph data distribution shifts, TSA employs three synergistic strategies: uncertainty-aware neighborhood weighting to accommodate neighbor label distribution shifts, adaptive balancing of self-node and aggregated neighborhood representations based on their signal-to-noise ratio, and decision boundary refinement to correct residual label and feature shifts. Extensive experiments on synthetic and real-world datasets demonstrate TSA’s consistent outperformance of both non-graph TTA methods and state-of-the-art GTTA baselines. } }
Endnote
%0 Conference Paper %T Structural Alignment Improves Graph Test-Time Adaptation %A Hans Hao-Hsun Hsu %A Shikun Liu %A Han Zhao %A Pan Li %B Proceedings of The 29th International Conference on Artificial Intelligence and Statistics %C Proceedings of Machine Learning Research %D 2026 %E Emtiyaz Khan %E Yingzhen Li %E Arno Solin %E Aaditya Ramdas %F pmlr-v300-hsu26a %I PMLR %P 829--837 %U https://proceedings.mlr.press/v300/hsu26a.html %V 300 %X Graph-based learning excels at capturing interaction patterns in diverse domains like recommendation, fraud detection, and particle physics. However, its performance often degrades under distribution shifts, especially those altering network connectivity. Current methods to address these shifts typically require retraining with the source dataset, which is often infeasible due to computational or privacy limitations. We introduce Test-Time Structural Alignment (TSA), a novel algorithm for Graph Test-Time Adaptation (GTTA) that adapts a pretrained model to align graph structures during inference without the cost of retraining. Grounded in a theoretical understanding of graph data distribution shifts, TSA employs three synergistic strategies: uncertainty-aware neighborhood weighting to accommodate neighbor label distribution shifts, adaptive balancing of self-node and aggregated neighborhood representations based on their signal-to-noise ratio, and decision boundary refinement to correct residual label and feature shifts. Extensive experiments on synthetic and real-world datasets demonstrate TSA’s consistent outperformance of both non-graph TTA methods and state-of-the-art GTTA baselines.
APA
Hsu, H.H., Liu, S., Zhao, H. & Li, P.. (2026). Structural Alignment Improves Graph Test-Time Adaptation . Proceedings of The 29th International Conference on Artificial Intelligence and Statistics, in Proceedings of Machine Learning Research 300:829-837 Available from https://proceedings.mlr.press/v300/hsu26a.html.

Related Material