arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2407.20899cs.AIcs.CL
收藏

面向图像分类的忠实且合理的自然语言解释:一种流水线方法

Faithful and Plausible Natural Language Explanations for Image Classification: A Pipeline Approach

Poznan University of Technology(波兹南理工大学);Charles University(查理大学);Samsung AI Center Warsaw(三星人工智能中心华沙分部)

Adam Wojciechowski, Mateusz Lango, Ondrej Dusek

AI总结:

本文提出一种事后自然语言解释流水线,通过分析关键神经元与激活图生成结构化意义表示并转为文本,在不改变 CNN 分类器训练和性能的前提下提升解释的忠实性与合理性。

摘要:

现有的图像分类解释方法难以同时提供忠实且合理的解释。本文针对这一问题,提出一种事后自然语言解释方法,该方法可应用于任何基于 CNN 的分类器,而无需改变其训练过程或影响预测性能。通过分析具有影响力的神经元及相应的激活图,该方法以结构化意义表示的形式生成对分类器决策过程的忠实描述,随后由语言模型将其转换为文本。通过这种流水线方法,所生成的解释植根于神经网络架构,能够提供对分类过程的准确洞察,同时仍便于非专家理解。实验结果表明,由本文方法构建的 NLE 在合理性和忠实性方面均显著更强。特别是,用户对神经网络结构的干预(神经元掩码)效果是基线方法的三倍。

英文摘要:

Existing explanation methods for image classification struggle to provide faithful and plausible explanations. This paper addresses this issue by proposing a post-hoc natural language explanation method that can be applied to any CNN-based classifier without altering its training process or affecting predictive performance. By analysing influential neurons and the corresponding activation maps, the method generates a faithful description of the classifier's decision process in the form of a structured meaning representation, which is then converted into text by a language model. Through this pipeline approach, the generated explanations are grounded in the neural network architecture, providing accurate insight into the classification process while remaining accessible to non-experts. Experimental results show that the NLEs constructed by our method are significantly more plausible and faithful. In particular, user interventions in the neural network structure (masking of neurons) are three times more effective than the baselines.

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。

↑