Spiking Graph Predictive Coding for Reliable OOD Generalization

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Ren, Jing, Du, Jiapeng, Li, Bowen, Xu, Ziqi, Zheng, Xin, Jia, Hong, Ma, Suyu, Xu, Xiwei, Xia, Feng
Format: Preprint
Published: 2026
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912919745200128
author Ren, Jing
Du, Jiapeng
Li, Bowen
Xu, Ziqi
Zheng, Xin
Jia, Hong
Ma, Suyu
Xu, Xiwei
Xia, Feng
author_facet Ren, Jing
Du, Jiapeng
Li, Bowen
Xu, Ziqi
Zheng, Xin
Jia, Hong
Ma, Suyu
Xu, Xiwei
Xia, Feng
contents Graphs provide a powerful basis for modeling Web-based relational data, with expressive GNNs to support the effective learning in dynamic web environments. However, real-world deployment is hindered by pervasive out-of-distribution (OOD) shifts, where evolving user activity and changing content semantics alter feature distributions and labeling criteria. These shifts often lead to unstable or overconfident predictions, undermining the trustworthiness required for Web4Good applications. Achieving reliable OOD generalization demands principled and interpretable uncertainty estimation; however, existing methods are largely post-hoc, insensitive to distribution shifts, and unable to explain where uncertainty arises especially in high-stakes settings. To address these limitations, we introduce SpIking GrapH predicTive coding (SIGHT), an uncertainty-aware plug-in graph learning module for reliable OOD Generalization. SIGHT performs iterative, error-driven correction over spiking graph states, enabling models to expose internal mismatch signals that reveal where predictions become unreliable. Across multiple graph benchmarks and diverse OOD scenarios, SIGHT consistently enhances predictive accuracy, uncertainty estimation, and interpretability when integrated with GNNs.
format Preprint
id arxiv_https___arxiv_org_abs_2602_19392
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Spiking Graph Predictive Coding for Reliable OOD Generalization
Ren, Jing
Du, Jiapeng
Li, Bowen
Xu, Ziqi
Zheng, Xin
Jia, Hong
Ma, Suyu
Xu, Xiwei
Xia, Feng
Machine Learning
Social and Information Networks
Graphs provide a powerful basis for modeling Web-based relational data, with expressive GNNs to support the effective learning in dynamic web environments. However, real-world deployment is hindered by pervasive out-of-distribution (OOD) shifts, where evolving user activity and changing content semantics alter feature distributions and labeling criteria. These shifts often lead to unstable or overconfident predictions, undermining the trustworthiness required for Web4Good applications. Achieving reliable OOD generalization demands principled and interpretable uncertainty estimation; however, existing methods are largely post-hoc, insensitive to distribution shifts, and unable to explain where uncertainty arises especially in high-stakes settings. To address these limitations, we introduce SpIking GrapH predicTive coding (SIGHT), an uncertainty-aware plug-in graph learning module for reliable OOD Generalization. SIGHT performs iterative, error-driven correction over spiking graph states, enabling models to expose internal mismatch signals that reveal where predictions become unreliable. Across multiple graph benchmarks and diverse OOD scenarios, SIGHT consistently enhances predictive accuracy, uncertainty estimation, and interpretability when integrated with GNNs.
title Spiking Graph Predictive Coding for Reliable OOD Generalization
topic Machine Learning
Social and Information Networks
url https://arxiv.org/abs/2602.19392