Graph Out-of-Distribution Detection via Test-Time Calibration with Dual Dynamic Dictionaries

TL;DR

This paper introduces BaCa, a test-time calibration method for graph out-of-distribution detection that uses dual dynamic dictionaries and a mix-up strategy to improve detection accuracy without fine-tuning.

Contribution

BaCa is a novel test-time graph OOD detection approach that calibrates scores with dual dictionaries and a mix-up strategy, avoiding fine-tuning and auxiliary datasets.

Findings

BaCa outperforms existing methods on real-world datasets.

It effectively captures latent ID and OOD representations.

The method improves boundary-aware OOD detection accuracy.

Abstract

A key challenge in graph out-of-distribution (OOD) detection lies in the absence of ground-truth OOD samples during training. Existing methods are typically optimized to capture features within the in-distribution (ID) data and calculate OOD scores, which often limits pre-trained models from representing distributional boundaries, leading to unreliable OOD detection. Moreover, the latent structure of graph data is often governed by multiple underlying factors, which remains less explored. To address these challenges, we propose a novel test-time graph OOD detection method, termed BaCa, that calibrates OOD scores using dual dynamically updated dictionaries without requiring fine-tuning the pre-trained model. Specifically, BaCa estimates graphons and applies a mix-up strategy solely with test samples to generate diverse boundary-aware discriminative topologies, eliminating the need for exposing auxiliary datasets as outliers. We construct dual dynamic dictionaries via priority queues and attention mechanisms to adaptively capture latent ID and OOD representations, which are then utilized for boundary-aware OOD score calibration. To the best of our knowledge, extensive experiments on real-world datasets show that BaCa significantly outperforms existing state-of-the-art methods in OOD detection.