Learning to Faithfully Rationalize by Construction
Learning to Faithfully Rationalize by Construction
复制标题
DOI：
10.18653/v1/2020.acl-main.409
复制
发表时间：
2020-04
期刊：
影响因子：
--
通讯作者：
Sarthak Jain;Sarah Wiegreffe;Yuval Pinter;Byron C. Wallace
中科院分区：
文献类型：
--
作者：
Sarthak Jain;Sarah Wiegreffe;Yuval Pinter;Byron C. Wallace
在许多设置中，重要的是要了解模型在NLP中进行特定的预测。在某些环境中，忠实的解释是一个忠实的解释，可以确保Lei等人（2016年）提出了一个模型，以确定神经文本分类的忠实理性。通过这种方法进行的离散输入代币会使训练复杂化，导致较高的差异，并需要仔细的超参数调整。例如，来自训练的模型的梯度用于诱导令牌输入的二进制标签，然后将其训练以预测独立的分类器模块。如果分类器在自动评估和手动评估中是任意复杂的，我们发现这种简单的框架的变体会产生优于“端到端”方法的预测性能，而在HTTPS上则更加通用和更易于训练。 //github.com/successar/fresh。
In many settings it is important for one to be able to understand why a model made a particular prediction. In NLP this often entails extracting snippets of an input text ‘responsible for’ corresponding model output; when such a snippet comprises tokens that indeed informed the model’s prediction, it is a faithful explanation. In some settings, faithfulness may be critical to ensure transparency. Lei et al. (2016) proposed a model to produce faithful rationales for neural text classification by defining independent snippet extraction and prediction modules. However, the discrete selection over input tokens performed by this method complicates training, leading to high variance and requiring careful hyperparameter tuning. We propose a simpler variant of this approach that provides faithful explanations by construction. In our scheme, named FRESH, arbitrary feature importance scores (e.g., gradients from a trained model) are used to induce binary labels over token inputs, which an extractor can be trained to predict. An independent classifier module is then trained exclusively on snippets provided by the extractor; these snippets thus constitute faithful explanations, even if the classifier is arbitrarily complex. In both automatic and manual evaluations we find that variants of this simple framework yield predictive performance superior to ‘end-to-end’ approaches, while being more general and easier to train. Code is available at https://github.com/successar/FRESH.