通过方向导数衡量文本解释对预测的影响，提升模型决策透明度。
Zero-Shot Faithful Textual Explanations via Directional-Derivative Influence on Predictions
- 用方向导数计算文本解释对分类结果的直接影响，评估其重要性。
- 在多个数据集上显著优于基线方法，解释更贴近真实决策依据。
- 适合关注模型可解释性与公平性的研究人员和开发者。
零样本文本解释旨在不依赖特定任务标注或大视觉语言模型（LVLMs）的情况下，通过探测图像分类器内部表示来增强模型透明性。然而，现有方法常遗漏真正驱动预测的关键特征，导致解释的忠实度不足。为此，我们提出FaithTrace。受忠实解释应描述对预测有显著影响概念的启发，FaithTrace直接测量由解释生成的表示对类别对数几率（logit）的影响。我们引入一种影响力评分，基于分类器特征空间中由文本诱导方向上的类别对数几率的方向导数，并将其作为忠实度的代理指标。此外，我们将该评分扩展为可量化的评估指标，填补了文本解释忠实度评估的空白。实验表明，FaithTrace生成的解释比基线方法更具忠实性，有助于更准确理解模型决策过程。代码将公开发布。
原文摘要 · Abstract (English)
Zero-shot textual explanations aim to make image classifiers more transparent by probing their internal representations, without relying on task-specific supervision or LVLMs. However, existing methods often miss the features that truly drive the prediction, resulting in limited \textit{faithfulness} to the evidence underlying the model's decision. To address this, we propose FaithTrace. Motivated by the idea that faithful explanations should describe concepts that strongly influence the prediction, FaithTrace directly measures how much the representation induced by the explanation changes the class logit. We introduce an influence score, computed as the directional derivative of the class logit along the text-induced direction in the classifier's feature space, and use it as a proxy for faithfulness. Moreover, we extend this influence score into quantitative evaluation metrics, helping fill the gap in faithfulness evaluation for textual explanations. Experiments show that FaithTrace yields more faithful explanations than baselines, facilitating a more accurate understanding of the model. The code will be publicly released.
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。