Research

Paper

TESTING February 19, 2026

From Subtle to Significant: Prompt-Driven Self-Improving Optimization in Test-Time Graph OOD Detection

Authors

Luzhi Wang, Xuanshuo Fu, He Zhang, Chuang Liu, Xiaobao Wang, Hongbo Liu

Abstract

Graph Out-of-Distribution (OOD) detection aims to identify whether a test graph deviates from the distribution of graphs observed during training, which is critical for ensuring the reliability of Graph Neural Networks (GNNs) when deployed in open-world scenarios. Recent advances in graph OOD detection have focused on test-time training techniques that facilitate OOD detection without accessing potential supervisory information (e.g., training data). However, most of these methods employ a one-pass inference paradigm, which prevents them from progressively correcting erroneous predictions to amplify OOD signals. To this end, we propose a \textbf{S}elf-\textbf{I}mproving \textbf{G}raph \textbf{O}ut-\textbf{o}f-\textbf{D}istribution detector (SIGOOD), which is an unsupervised framework that integrates continuous self-learning with test-time training for effective graph OOD detection. Specifically, SIGOOD generates a prompt to construct a prompt-enhanced graph that amplifies potential OOD signals. To optimize prompts, SIGOOD introduces an Energy Preference Optimization (EPO) loss, which leverages energy variations between the original test graph and the prompt-enhanced graph. By iteratively optimizing the prompt by involving it into the detection model in a self-improving loop, the resulting optimal prompt-enhanced graph is ultimately used for OOD detection. Comprehensive evaluations on 21 real-world datasets confirm the effectiveness and outperformance of our SIGOOD method. The code is at https://github.com/Ee1s/SIGOOD.

Metadata

arXiv ID: 2602.17342
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-02-19
Fetched: 2026-02-21 18:51

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2602.17342v1</id>\n    <title>From Subtle to Significant: Prompt-Driven Self-Improving Optimization in Test-Time Graph OOD Detection</title>\n    <updated>2026-02-19T13:19:53Z</updated>\n    <link href='https://arxiv.org/abs/2602.17342v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2602.17342v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Graph Out-of-Distribution (OOD) detection aims to identify whether a test graph deviates from the distribution of graphs observed during training, which is critical for ensuring the reliability of Graph Neural Networks (GNNs) when deployed in open-world scenarios. Recent advances in graph OOD detection have focused on test-time training techniques that facilitate OOD detection without accessing potential supervisory information (e.g., training data). However, most of these methods employ a one-pass inference paradigm, which prevents them from progressively correcting erroneous predictions to amplify OOD signals. To this end, we propose a \\textbf{S}elf-\\textbf{I}mproving \\textbf{G}raph \\textbf{O}ut-\\textbf{o}f-\\textbf{D}istribution detector (SIGOOD), which is an unsupervised framework that integrates continuous self-learning with test-time training for effective graph OOD detection. Specifically, SIGOOD generates a prompt to construct a prompt-enhanced graph that amplifies potential OOD signals. To optimize prompts, SIGOOD introduces an Energy Preference Optimization (EPO) loss, which leverages energy variations between the original test graph and the prompt-enhanced graph. By iteratively optimizing the prompt by involving it into the detection model in a self-improving loop, the resulting optimal prompt-enhanced graph is ultimately used for OOD detection. Comprehensive evaluations on 21 real-world datasets confirm the effectiveness and outperformance of our SIGOOD method. The code is at https://github.com/Ee1s/SIGOOD.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.AI'/>\n    <published>2026-02-19T13:19:53Z</published>\n    <arxiv:comment>9pages, 5 figures</arxiv:comment>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Luzhi Wang</name>\n    </author>\n    <author>\n      <name>Xuanshuo Fu</name>\n    </author>\n    <author>\n      <name>He Zhang</name>\n    </author>\n    <author>\n      <name>Chuang Liu</name>\n    </author>\n    <author>\n      <name>Xiaobao Wang</name>\n    </author>\n    <author>\n      <name>Hongbo Liu</name>\n    </author>\n  </entry>"
}