GAIA: Delving into Gradient-based Attribution Abnormality for Out-of-distribution Detection

TL;DR

GAIA leverages abnormalities in gradient-based attribution explanations to effectively detect out-of-distribution data, improving reliability of neural networks in real-world applications.

Contribution

The paper introduces GAIA, a novel method that analyzes attribution gradient abnormalities for OOD detection, addressing challenges faced by existing explanation-based approaches.

Findings

GAIA reduces FPR95 by 23.10% on CIFAR10

GAIA reduces FPR95 by 45.41% on CIFAR100

Effective on both CIFAR and ImageNet-1k benchmarks

Abstract

Detecting out-of-distribution (OOD) examples is crucial to guarantee the reliability and safety of deep neural networks in real-world settings. In this paper, we offer an innovative perspective on quantifying the disparities between in-distribution (ID) and OOD data -- analyzing the uncertainty that arises when models attempt to explain their predictive decisions. This perspective is motivated by our observation that gradient-based attribution methods encounter challenges in assigning feature importance to OOD data, thereby yielding divergent explanation patterns. Consequently, we investigate how attribution gradients lead to uncertain explanation outcomes and introduce two forms of abnormalities for OOD detection: the zero-deflation abnormality and the channel-wise average abnormality. We then propose GAIA, a simple and effective approach that incorporates Gradient Abnormality Inspection and Aggregation. The effectiveness of GAIA is validated on both commonly utilized (CIFAR) and large-scale (ImageNet-1k) benchmarks. Specifically, GAIA reduces the average FPR95 by 23.10% on CIFAR10 and by 45.41% on CIFAR100 compared to advanced post-hoc methods.