Knowledge Commons of Institute of Automation,CAS
Logic Traps in Evaluating Attribution Scores | |
Ju YM(鞠一鸣)![]() ![]() ![]() ![]() | |
2022-05 | |
会议名称 | Proceedings of the 60th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers) |
页码 | 5911–5922 |
会议日期 | 22nd - 27th May 2022 |
会议地点 | Dublin |
摘要 | Modern deep learning models are notoriously opaque, which has motivated the development of methods for interpreting how deep models predict.This goal is usually approached with attribution method, which assesses the influence of features on model predictions. As an explanation method, the evaluation criteria of attribution methods is how accurately it reflects the actual reasoning process of the model (faithfulness). Meanwhile, since the reasoning process of deep models is inaccessible, researchers design various evaluation methods to demonstrate their arguments.However, some crucial logic traps in these evaluation methods are ignored in most works, causing inaccurate evaluation and unfair comparison.This paper systematically reviews existing methods for evaluating attribution scores and summarizes the logic traps in these methods.We further conduct experiments to demonstrate the existence of each logic trap.Through both theoretical and experimental analysis, we hope to increase attention on the inaccurate evaluation of attribution scores. Moreover, with this paper, we suggest stopping focusing on improving performance under unreliable evaluation systems and starting efforts on reducing the impact of proposed logic traps. |
收录类别 | EI |
语种 | 英语 |
是否为代表性论文 | 是 |
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 | 可解释人工智能 |
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/173211/52277 |
专题 | 多模态人工智能系统全国重点实验室_自然语言处理 |
作者单位 | 1.National Laboratory of Pattern Recognition, Institute of Automation, CAS, Beijing, China 2.School of Artificial Intelligence, University of Chinese Academy of Sciences, Beijing, China |
推荐引用方式 GB/T 7714 | Ju YM,Zhang YZ,Yang C,et al. Logic Traps in Evaluating Attribution Scores[C],2022:5911–5922. |
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 | ||
2022.acl-long.407 (1(1073KB) | 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