arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2106.00786cs.LGcs.AIcs.CL
收藏

特征重要性解释中的分布外问题与搜索方法

The Out-of-Distribution Problem in Explainability and Search Methods for Feature Importance Explanations

University of North Carolina at Chapel Hill(北卡罗来纳大学教堂山分校)

Peter Hase, Harry Xie, Mohit Bansal

摘要:

特征重要性(FI)估计是一种流行的解释形式,通常通过计算在测试时删除某些输入特征引起的模型置信度变化来创建和评估。例如,在标准的Sufficiency度量中,仅保留前k个最重要的令牌。在本文中,我们研究了几个不足的FI解释的维度,这种形式的解释提供概念和实证的改进。首先,我们提出了一个新的论点,为什么在创建或评估解释时从输入中删除特征会有问题:这些反事实输入对模型来说是分布外的(OOD),这意味着由此产生的解释在社会上是不一致的。问题的关键在于模型先验和随机权重初始化以意想不到的方式影响解释(和解释度量)。为了解决这个问题,我们提出了一个简单的修改模型训练过程,这将导致更多的社会一致的解释和指标。其次,我们比较了从模型输入中去除特征的五种方法。我们发现,一些方法比其他方法产生更多的OOD反事实,我们提出建议,选择一个功能替换功能。最后,我们介绍了四种基于搜索的方法,用于识别FI解释,并将它们与强基线进行比较,包括LIME,Covereors和Integrated Covereents。通过六个不同的文本分类数据集的实验,我们发现,唯一的方法,始终优于随机搜索是一个并行本地搜索(PLS),我们介绍。与第二好的方法相比,充分性提高了5.4分,全面性提高了17分。本文中所有支持实验的代码都可以在https://github.com/peterbhase/ExplanationSearch上公开获得。

英文摘要:

Feature importance (FI) estimates are a popular form of explanation, and they are commonly created and evaluated by computing the change in model confidence caused by removing certain input features at test time. For example, in the standard Sufficiency metric, only the top-k most important tokens are kept. In this paper, we study several under-explored dimensions of FI explanations, providing conceptual and empirical improvements for this form of explanation. First, we advance a new argument for why it can be problematic to remove features from an input when creating or evaluating explanations: the fact that these counterfactual inputs are out-of-distribution (OOD) to models implies that the resulting explanations are socially misaligned. The crux of the problem is that the model prior and random weight initialization influence the explanations (and explanation metrics) in unintended ways. To resolve this issue, we propose a simple alteration to the model training process, which results in more socially aligned explanations and metrics. Second, we compare among five approaches for removing features from model inputs. We find that some methods produce more OOD counterfactuals than others, and we make recommendations for selecting a feature-replacement function. Finally, we introduce four search-based methods for identifying FI explanations and compare them to strong baselines, including LIME, Anchors, and Integrated Gradients. Through experiments with six diverse text classification datasets, we find that the only method that consistently outperforms random search is a Parallel Local Search (PLS) that we introduce. Improvements over the second-best method are as large as 5.4 points for Sufficiency and 17 points for Comprehensiveness. All supporting code for experiments in this paper is publicly available at https://github.com/peterbhase/ExplanationSearch.

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。

↑