Knowledge Commons of Institute of Automation,CAS
Can We Really Trust Explanations? Evaluating the Stability of Feature Attribution Explanation Methods via Adversarial Attack | |
Zhao Yang1,2; Yuanzhe Zhang1,2; Zhongtao Jiang1,2; Yiming Ju1,2 | |
2022 | |
会议名称 | Proceedings of the 21st Chinese National Conference on Computational Linguistics |
会议日期 | 2022-10 |
会议地点 | Nanchang |
摘要 | Explanations can increase the transparency of neural networks and make them more trustworthy. However, can we really trust explanations generated by the existing explanation methods? If the explanation methods are not stable enough, the credibility of the explanation will be greatly reduced. Previous studies seldom considered such an important issue. To this end, this paper proposes a new evaluation frame to evaluate the stability of current typical feature attribution explanation methods via textual adversarial attack. Our frame could generate adversarial examples with similar textual semantics. Such adversarial examples will make the original models have the same outputs, but make most current explanation methods deduce completely different explanations. Under this frame, we test five classical explanation methods and show their performance on several stability-related metrics. Experimental results show our evaluation is effective and could reveal the stability performance of existing explanation methods. |
收录类别 | EI |
七大方向——子方向分类 | 自然语言处理 |
国重实验室规划方向分类 | 语音语言处理 |
是否有论文关联数据集需要存交 | 否 |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/56725 |
专题 | 复杂系统认知与决策实验室 |
作者单位 | 1.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China 2.National Laboratory of Pattern Recognition, Institute of Automation, CAS, Beijing, China |
第一作者单位 | 模式识别国家重点实验室 |
推荐引用方式 GB/T 7714 | Zhao Yang,Yuanzhe Zhang,Zhongtao Jiang,et al. Can We Really Trust Explanations? Evaluating the Stability of Feature Attribution Explanation Methods via Adversarial Attack[C],2022. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
杨朝-CCL.pdf(355KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论