Large language models are increasingly capable of generating fluent-appearing text with relatively little task-specific supervision. But can these models accurately explain classification decisions? We consider the task of generating free-text explanations using a small number of human-written examples (i.e., in a few-shot manner). We find that (1) authoring higher-quality examples for prompting results in higher quality generations; and (2) surprisingly, in a head-to-head comparison, crowdworkers often prefer explanations generated by GPT-3 to crowdsourced human-written explanations contained within existing datasets. Crowdworker ratings also show, however, that while models produce factual, grammatical, and sufficient explanations, they have room to improve, e.g., along axes such as providing novel information and supporting the label. We create a pipeline that combines GPT-3 with a supervised filter that incorporates humans-in-the-loop via binary acceptability judgments. Despite significant subjectivity intrinsic to judging acceptability, our approach is able to consistently filter GPT-3 generated explanations deemed acceptable by humans.
The perceived toxicity of language can vary based on someone's identity and beliefs, but this variation is often ignored when collecting toxic language datasets, resulting in dataset and model biases. We seek to understand the who, why, and what behind biases in toxicity annotations. In two online studies with demographically and politically diverse participants, we investigate the effect of annotator identities (who) and beliefs (why), drawing from social psychology research about hate speech, free speech, racist beliefs, political leaning, and more. We disentangle what is annotated as toxic by considering posts with three characteristics: anti-Black language, African American English (AAE) dialect, and vulgarity. Our results show strong associations between annotator identity and beliefs and their ratings of toxicity. Notably, more conservative annotators and those who scored highly on our scale for racist beliefs were less likely to rate anti-Black language as toxic, but more likely to rate AAE as toxic. We additionally present a case study illustrating how a popular toxicity detection system's ratings inherently reflect only specific beliefs and perspectives. Our findings call for contextualizing toxicity labels in social variables, which raises immense implications for toxic language annotation and detection.
Estimating the difficulty of a dataset typically involves comparing state-of-the-art models to humans; the bigger the performance gap, the harder the dataset is said to be. Not only is this framework informal, but it also provides little understanding of how difficult each instance is, or what attributes make it difficult for a given model. To address these problems, we propose an information-theoretic perspective, framing dataset difficulty as the absence of $\textit{usable information}$. Measuring usable information is as easy as measuring performance, but has certain theoretical advantages. While the latter only allows us to compare different models w.r.t the same dataset, the former also allows us to compare different datasets w.r.t the same model. We then introduce $\textit{pointwise}$ $\mathcal{V}-$$\textit{information}$ (PVI) for measuring the difficulty of individual instances, where instances with higher PVI are easier for model $\mathcal{V}$. By manipulating the input before measuring usable information, we can understand $\textit{why}$ a dataset is easy or difficult for a given model, which we use to discover annotation artefacts in widely-used benchmarks.
While FrameNet is widely regarded as a rich resource of semantics in natural language processing, a major criticism concerns its lack of coverage and the relative paucity of its labeled data compared to other commonly used lexical resources such as PropBank and VerbNet. This paper reports on a pilot study to address these gaps. We propose a data augmentation approach, which uses existing frame-specific annotation to automatically annotate other lexical units of the same frame which are unannotated. Our rule-based approach defines the notion of a sister lexical unit and generates frame-specific augmented data for training. We present experiments on frame-semantic role labeling which demonstrate the importance of this data augmentation: we obtain a large improvement to prior results on frame identification and argument identification for FrameNet, utilizing both full-text and lexicographic annotations under FrameNet. Our findings on data augmentation highlight the value of automatic resource creation for improved models in frame-semantic parsing.
Despite recent advances in natural language generation, it remains challenging to control attributes of generated text. We propose DExperts: Decoding-time Experts, a decoding-time method for controlled text generation that combines a pretrained language model with "expert" LMs and/or "anti-expert" LMs in a product of experts. Intuitively, under the ensemble, tokens only get high probability if they are considered likely by the experts, and unlikely by the anti-experts. We apply DExperts to language detoxification and sentiment-controlled generation, where we outperform existing controllable generation methods on both automatic and human evaluations. Moreover, because DExperts operates only on the output of the pretrained LM, it is effective with (anti-)experts of smaller size, including when operating on GPT-3. Our work highlights the promise of tuning small LMs on text with (un)desirable attributes for efficient decoding-time steering.
Despite recent advances in natural language generation, it remains challenging to control attributes of generated text. We propose DExperts: Decoding-time Experts, a decoding-time method for controlled text generation which combines a pretrained language model with experts and/or anti-experts in an ensemble of language models. Intuitively, under our ensemble, output tokens only get high probability if they are considered likely by the experts, and unlikely by the anti-experts. We apply DExperts to language detoxification and sentiment-controlled generation, where we outperform existing controllable generation methods on both automatic and human evaluations. Our work highlights the promise of using LMs trained on text with (un)desired attributes for efficient decoding-time controlled language generation.
Contrastive explanations clarify why an event occurred in contrast to another. They are more inherently intuitive to humans to both produce and comprehend. We propose a methodology to produce contrastive explanations for classification models by modifying the representation to disregard non-contrastive information, and modifying model behavior to only be based on contrastive reasoning. Our method is based on projecting model representation to a latent space that captures only the features that are useful (to the model) to differentiate two potential decisions. We demonstrate the value of contrastive explanations by analyzing two different scenarios, using both high-level abstract concept attribution and low-level input token/span attribution, on two widely used text classification tasks. Specifically, we produce explanations for answering: for which label, and against which alternative label, is some aspect of the input useful? And which aspects of the input are useful for and against particular decisions? Overall, our findings shed light on the ability of label-contrastive explanations to provide a more accurate and finer-grained interpretability of a model's decision.
Despite major advances in open-ended text generation, there has been limited progress in designing evaluation metrics for this task. We propose MAUVE -- a metric for open-ended text generation, which directly compares the distribution of machine-generated text to that of human language. MAUVE measures the mean area under the divergence curve for the two distributions, exploring the trade-off between two types of errors: those arising from parts of the human distribution that the model distribution approximates well, and those it does not. We present experiments across two open-ended generation tasks in the web text domain and the story domain, and a variety of decoding algorithms and model sizes. Our results show that evaluation under MAUVE indeed reflects the more natural behavior with respect to model size, compared to prior metrics. MAUVE's ordering of the decoding algorithms also agrees with that of generation perplexity, the most widely used metric in open-ended text generation; however, MAUVE presents a more principled evaluation metric for the task as it considers both model and human text.
Biased associations have been a challenge in the development of classifiers for detecting toxic language, hindering both fairness and accuracy. As potential solutions, we investigate recently introduced debiasing methods for text classification datasets and models, as applied to toxic language detection. Our focus is on lexical (e.g., swear words, slurs, identity mentions) and dialectal markers (specifically African American English). Our comprehensive experiments establish that existing methods are limited in their ability to prevent biased behavior in current toxicity detectors. We then propose an automatic, dialect-aware data correction method, as a proof-of-concept. Despite the use of synthetic labels, this method reduces dialectal associations with toxicity. Overall, our findings show that debiasing a model trained on biased toxic language data is not as effective as simply relabeling the data to remove existing biases.
Large datasets have become commonplace in NLP research. However, the increased emphasis on data quantity has made it challenging to assess the quality of data. We introduce Data Maps---a model-based tool to characterize and diagnose datasets. We leverage a largely ignored source of information: the behavior of the model on individual instances during training (training dynamics) for building data maps. This yields two intuitive measures for each example---the model's confidence in the true class, and the variability of this confidence across epochs---obtained in a single run of training. Experiments across four datasets show that these model-dependent measures reveal three distinct regions in the data map, each with pronounced characteristics. First, our data maps show the presence of "ambiguous" regions with respect to the model, which contribute the most towards out-of-distribution generalization. Second, the most populous regions in the data are "easy to learn" for the model, and play an important role in model optimization. Finally, data maps uncover a region with instances that the model finds "hard to learn"; these often correspond to labeling errors. Our results indicate that a shift in focus from quantity to quality of data could lead to robust models and improved out-of-distribution generalization.