揭示梯度解释方法的谱偏差，提出改进方案。
On Spectral Properties of Gradient-based Explanation Methods
- 从概率与谱分析角度形式化解释方法
- 发现梯度导致普遍谱偏差，平方梯度与扰动设计可缓解
- 提出标准扰动尺度与SpectralLens聚合方法，适合模型可解释性研究者
理解深度网络行为对提升其结果可信度至关重要。尽管已有大量研究致力于解释预测，但可靠性问题仍存，根源在于缺乏充分的形式化。本文采用新颖的概率与谱视角，形式化分析解释方法。研究揭示了由梯度使用引发的普遍谱偏差，并解释了部分实验发现的设计选择，如使用平方梯度和输入扰动。进一步分析表明，解释方法中扰动超参数的选择可能导致不一致的解释，为此基于所提形式化框架提出两种修复方法：(i) 确定标准扰动尺度的机制，(ii) 称为SpectralLens的聚合方法。最后通过定量评估验证了理论结果。
原文摘要 · Abstract (English)
Understanding the behavior of deep networks is crucial to increase our confidence in their results. Despite an extensive body of work for explaining their predictions, researchers have faced reliability issues, which can be attributed to insufficient formalism. In our research, we adopt novel probabilistic and spectral perspectives to formally analyze explanation methods. Our study reveals a pervasive spectral bias stemming from the use of gradient, and sheds light on some common design choices that have been discovered experimentally, in particular, the use of squared gradient and input perturbation. We further characterize how the choice of perturbation hyperparameters in explanation methods, such as SmoothGrad, can lead to inconsistent explanations and introduce two remedies based on our proposed formalism: (i) a mechanism to determine a standard perturbation scale, and (ii) an aggregation method which we call SpectralLens. Finally, we substantiate our theoretical results through quantitative evaluations.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。