aiqbench analysis
This paper introduces N-GNAS, a novel framework that automates the selection of network architectures at the node level to address over-smoothing in Graph Neural Networks. By integrating contrastive learning, the method improves performance on non-Euclidean data compared to traditional uniform convolution approaches and existing GNAS techniques. The work represents a significant advancement in adaptive machine learning architectures for handling complex, unstructured datasets.
Node-level Graph Neural Architecture Search Framework
arXiv:2610.09297v1 Announce Type: new Abstract: In recent years, Graph Neural Networks (GNNs) and architecture search frameworks have gained extensive application in non-Euclidean data processing, attributable to their superior capacity in managing unstructured data. Nevertheless, traditional approaches typically apply uniform convolution operations to all nodes, regardless of their varying structural and feature characteristics, which can undermine model performance and result in over-smoothing issues as the number of layers increases. To overcome this limitation, in this work, we propose a \textbf{N}ode-Level \textbf{G}raph \textbf{N}eural \textbf{A}rchitecture \textbf{S}earch (N-GNAS) algorithm. It can automatically choose an appropriate network architecture for each subset of nodes when updating node features. N-GNAS also introduces a contrastive learning loss to separate sample features from different categories and vice versa. In experiments conducted on eight datasets for node and graph classification, our methodology outperforms current leading GNAS techniques and traditional human-designed GNNs. For example, it achieves an accuracy rate of 78.26\% on the CiteSeer dataset.
Abstract:In recent years, Graph Neural Networks (GNNs) and architecture search frameworks have gained extensive application in non-Euclidean data processing, attributable to their superior capacity in managing unstructured data. Nevertheless, traditional approaches typically apply uniform convolution operations to all nodes, regardless of their varying structural and feature characteristics, which can undermine model performance and result in over-smoothing issues as the number of layers increases. To overcome this limitation, in this work, we propose a \textbf{N}ode-Level \textbf{G}raph \textbf{N}eural \textbf{A}rchitecture \textbf{S}earch (N-GNAS) algorithm. It can automatically choose an appropriate network architecture for each subset of nodes when updating node features. N-GNAS also introduces a contrastive learning loss to separate sample features from different categories and vice versa. In experiments conducted on eight datasets for node and graph classification, our methodology outperforms current leading GNAS techniques and traditional human-designed GNNs. For example, it achieves an accuracy rate of 78.26\% on the CiteSeer dataset.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2610.09297 [cs.LG]
(or arXiv:2610.09297v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.09297 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 246 words
Related coverage
A Polynomial Architecture-Attribution Co-Design Framework for Exact Aumann-Shapley Attribution in GNNs
arXiv · Jul 24, 2026
Node-wise Feature Encoding for Neural Performance Prediction
arXiv · Aug 31, 2026
Learning Structural Manipulability in Gate-Level Netlists Using Graph Neural Networks
arXiv · Jul 21, 2026
EdgeRefine: Privacy-Utility Balance for Graphs via Jaccard Sampling under Edge Differential Privacy
arXiv · Jul 29, 2026