arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2504.15664cs.LGcs.CV
收藏

基于XAI的神经网络捷径学习分析

An XAI-based Analysis of Shortcut Learning in Neural Networks

University of Marburg(马尔堡大学);University of Mannheim(曼海姆大学)

Phuong Quynh Le, Jörg Schlötterer, Christin Seifert

AI总结:

本文提出基于XAI的神经元虚假分数,系统分析CNN和ViT中虚假特征的编码与解耦程度,揭示现有缓解方法假设不完整,为开发新方法奠定基础。

摘要:

机器学习模型倾向于学习虚假特征——即与目标标签强相关但并非因果关系的特征。现有的减轻模型对虚假特征依赖的方法在某些情况下有效,但在其他情况下失效。在本文中,我们系统地分析了神经网络如何以及在何处编码虚假相关性。我们引入了神经元虚假分数,一种基于XAI的诊断度量,用于量化神经元对虚假特征的依赖程度。我们使用架构特定的方法分析了卷积神经网络(CNN)和视觉变换器(ViT)。我们的结果表明,虚假特征部分地被解耦,但解耦程度因模型架构而异。此外,我们发现现有缓解方法背后的假设是不完整的。我们的结果为开发新的缓解虚假相关性的方法奠定了基础,并使AI模型在实践中更安全地使用。

英文摘要:

Machine learning models tend to learn spurious features - features that strongly correlate with target labels but are not causal. Existing approaches to mitigate models' dependence on spurious features work in some cases, but fail in others. In this paper, we systematically analyze how and where neural networks encode spurious correlations. We introduce the neuron spurious score, an XAI-based diagnostic measure to quantify a neuron's dependence on spurious features. We analyze both convolutional neural networks (CNNs) and vision transformers (ViTs) using architecture-specific methods. Our results show that spurious features are partially disentangled, but the degree of disentanglement varies across model architectures. Furthermore, we find that the assumptions behind existing mitigation methods are incomplete. Our results lay the groundwork for the development of novel methods to mitigate spurious correlations and make AI models safer to use in practice.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑