Despite the strong performance of current NLP models, they can be brittle against adversarial attacks. To enable effective learning against adversarial inputs, we introduce the use of rationale models that can explicitly learn to ignore attack tokens. We find that the rationale models can successfully ignore over 90% of attack tokens. This approach leads to consistent sizable improvements (~10%) over baseline models in robustness on three datasets for both BERT and RoBERTa, and also reliably outperforms data augmentation with adversarial examples alone. In many cases, we find that our method is able to close the gap between model performance on a clean test set and an attacked test set and hence reduce the effect of adversarial attacks.

Learning from rationales seeks to augment model prediction accuracy using human-annotated rationales (i.e. subsets of input tokens) that justify their chosen labels, often in the form of intermediate or multitask supervision. While intuitive, this idea has proven elusive in practice. We make two observations about human rationales via empirical analyses:1) maximizing rationale supervision accuracy is not necessarily the optimal objective for improving model accuracy; 2) human rationales vary in whether they provide sufficient information for the model to exploit for prediction. Building on these insights, we propose several novel loss functions and learning strategies, and evaluate their effectiveness on three datasets with human rationales. Our results demonstrate consistent improvements over baselines in both label and rationale accuracy, including a 3% accuracy improvement on MultiRC. Our work highlights the importance of understanding properties of human explanations and exploiting them accordingly in model training.

The NLP community are increasingly interested in providing explanations for NLP models to help people make sense of model behavior and potentially improve human interaction with models. In addition to computational challenges in generating these explanations, evaluations of the generated explanations require human-centered perspectives and approaches. This tutorial will provide an overview of human-centered evaluations of explanations. First, we will give a brief introduction to the psychological foundation of explanations as well as types of NLP model explanations and their corresponding presentation, to provide the necessary background. We will then present a taxonomy of human-centered evaluation of explanations and dive into depth in the two categories: 1) evaluation based on human-annotated explanations; 2) evaluation with human-subjects studies. We will conclude by discussing future directions. We will also adopt a flipped format to maximize the in- teractive components for the live audience.

Text simplification reduces the language complexity of professional content for accessibility purposes. End-to-end neural network models have been widely adopted to directly generate the simplified version of input text, usually functioning as a blackbox. We show that text simplification can be decomposed into a compact pipeline of tasks to ensure the transparency and explainability of the process. The first two steps in this pipeline are often neglected: 1) to predict whether a given piece of text needs to be simplified, and 2) if yes, to identify complex parts of the text. The two tasks can be solved separately using either lexical or deep learning methods, or solved jointly. Simply applying explainable complexity prediction as a preliminary step, the out-of-sample text simplification performance of the state-of-the-art, black-box simplification models can be improved by a large margin.

Two main approaches for evaluating the quality of machine-generated rationales are: 1) using human rationales as a gold standard; and 2) automated metrics based on how rationales affect model behavior. An open question, however, is how human rationales fare with these automatic metrics. Analyzing a variety of datasets and models, we find that human rationales do not necessarily perform well on these metrics. To unpack this finding, we propose improved metrics to account for model-dependent baseline performance. We then propose two methods to further characterize rationale quality, one based on model retraining and one on using “fidelity curves” to reveal properties such as irrelevance and redundancy. Our work leads to actionable suggestions for evaluating and characterizing rationales.

We conduct a large-scale, systematic study to evaluate the existing evaluation methods for natural language generation in the context of generating online product reviews. We compare human-based evaluators with a variety of automated evaluation procedures, including discriminative evaluators that measure how well machine-generated text can be distinguished from human-written text, as well as word overlap metrics that assess how similar the generated text compares to human-written references. We determine to what extent these different evaluators agree on the ranking of a dozen of state-of-the-art generators for online product reviews. We find that human evaluators do not correlate well with discriminative evaluators, leaving a bigger question of whether adversarial accuracy is the correct objective for natural language generation. In general, distinguishing machine-generated text is challenging even for human evaluators, and human decisions correlate better with lexical overlaps. We find lexical diversity an intriguing metric that is indicative of the assessments of different evaluators. A post-experiment survey of participants provides insights into how to evaluate and improve the quality of natural language generation systems.

We introduce an adversarial method for producing high-recall explanations of neural text classifier decisions. Building on an existing architecture for extractive explanations via hard attention, we add an adversarial layer which scans the residual of the attention for remaining predictive signal. Motivated by the important domain of detecting personal attacks in social media comments, we additionally demonstrate the importance of manually setting a semantically appropriate “default” behavior for the model by explicitly manipulating its bias term. We develop a validation set of human-annotated personal attacks to evaluate the impact of these changes.