Pair the Dots: Jointly Examining Training History and Test Stimuli for Model Interpretability

Any prediction from a model is made by a combination of learning history and test stimuli. This provides significant insights for improving model interpretability: {\it because of which part(s) of which training example(s), the model attends to which part(s) of a test example}. Unfortunately, existing methods to interpret a model's predictions are only able to capture a single aspect of either test stimuli or learning history, and evidences from both are never combined or integrated. In this paper, we propose an efficient and differentiable approach to make it feasible to interpret a model's prediction by jointly examining training history and test stimuli. Test stimuli is first identified by gradient-based methods, signifying {\it the part of a test example that the model attends to}. The gradient-based saliency scores are then propagated to training examples using influence functions to identify {\it which part(s) of which training example(s)} make the model attends to the test stimuli. The system is differentiable and time efficient: the adoption of saliency scores from gradient-based methods allows us to efficiently trace a model's prediction through test stimuli, and then back to training examples through influence functions. We demonstrate that the proposed methodology offers clear explanations about neural model decisions, along with being useful for performing error analysis, crafting adversarial examples and fixing erroneously classified examples.

Plan ahead: Self-Supervised Text Planning for Paragraph Completion Task

Despite the recent success of contextualized language models on various NLP tasks, language model itself cannot capture textual coherence of a long, multi-sentence document (e.g., a paragraph). Humans often make structural decisions on what and how to say about before making utterances. Guiding surface realization with such high-level decisions and structuring text in a coherent way is essentially called a planning process. Where can the model learn such high-level coherence? A paragraph itself contains various forms of inductive coherence signals called self-supervision in this work, such as sentence orders, topical keywords, rhetorical structures, and so on. Motivated by that, this work proposes a new paragraph completion task PARCOM; predicting masked sentences in a paragraph. However, the task suffers from predicting and selecting appropriate topical content with respect to the given context. To address that, we propose a self-supervised text planner SSPlanner that predicts what to say first (content prediction), then guides the pretrained language model (surface realization) using the predicted content. SSPlanner outperforms the baseline generation models on the paragraph completion task in both automatic and human evaluation. We also find that a combination of noun and verb types of keywords is the most effective for content selection. As more number of content keywords are provided, overall generation quality also increases.

GenAug: Data Augmentation for Finetuning Text Generators

In this paper, we investigate data augmentation for text generation, which we call GenAug. Text generation and language modeling are important tasks within natural language processing, and are especially challenging for low-data regimes. We propose and evaluate various augmentation methods, including some that incorporate external knowledge, for finetuning GPT-2 on a subset of Yelp Reviews. We also examine the relationship between the amount of augmentation and the quality of the generated text. We utilize several metrics that evaluate important aspects of the generated text including its diversity and fluency. Our experiments demonstrate that insertion of character-level synthetic noise and keyword replacement with hypernyms are effective augmentation methods, and that the quality of generations improves to a peak at approximately three times the amount of original data.

BERTering RAMS: What and How Much does BERT Already Know About Event Arguments? -- A Study on the RAMS Dataset

Using the attention map based probing frame-work from (Clark et al., 2019), we observe that, on the RAMS dataset (Ebner et al., 2020), BERT's attention heads have modest but well above-chance ability to spot event arguments sans any training or domain finetuning, vary-ing from a low of 17.77% for Place to a high of 51.61% for Artifact. Next, we find that linear combinations of these heads, estimated with approx 11% of available total event argument detection supervision, can push performance well-higher for some roles - highest two being Victim (68.29% Accuracy) and Artifact(58.82% Accuracy). Furthermore, we investigate how well our methods do for cross-sentence event arguments. We propose a procedure to isolate "best heads" for cross-sentence argument detection separately of those for intra-sentence arguments. The heads thus estimated have superior cross-sentence performance compared to their jointly estimated equivalents, albeit only under the unrealistic assumption that we already know the argument is present in an-other sentence. Lastly, we seek to isolate to what extent our numbers stem from lexical frequency based associations between gold arguments and roles. We propose NONCE, a scheme to create adversarial test examples by replacing gold arguments with randomly generated "nonce" words. We find that learnt linear combinations are robust to NONCE, though individual best heads can be more sensitive.

Detecting Attackable Sentences in Arguments

Finding attackable sentences in an argument is the first step toward successful refutation in argumentation. We present a first large-scale analysis of sentence attackability in online arguments. We analyze driving reasons for attacks in argumentation and identify relevant characteristics of sentences. We demonstrate that a sentence's attackability is associated with many of these characteristics regarding the sentence's content, proposition types, and tone, and that an external knowledge source can provide useful information about attackability. Building on these findings, we demonstrate that machine learning models can automatically detect attackable sentences in arguments, significantly better than several baselines and comparably well to laypeople.

Extracting Implicitly Asserted Propositions in Argumentation

Argumentation accommodates various rhetorical devices, such as questions, reported speech, and imperatives. These rhetorical tools usually assert argumentatively relevant propositions rather implicitly, so understanding their true meaning is key to understanding certain arguments properly. However, most argument mining systems and computational linguistics research have paid little attention to implicitly asserted propositions in argumentation. In this paper, we examine a wide range of computational methods for extracting propositions that are implicitly asserted in questions, reported speech, and imperatives in argumentation. By evaluating the models on a corpus of 2016 U.S. presidential debates and online commentary, we demonstrate the effectiveness and limitations of the computational models. Our study may inform future research on argument mining and the semantics of these rhetorical devices in argumentation.

Explaining The Efficacy of Counterfactually-Augmented Data

In attempts to produce machine learning models less reliant on spurious patterns in training data, researchers have recently proposed a human-in-the-loop process for generating counterfactually augmented datasets. As applied in NLP, given some documents and their (initial) labels, humans are tasked with revising the text to make a (given) counterfactual label applicable. Importantly, the instructions prohibit edits that are not necessary to flip the applicable label. Models trained on the augmented (original and revised) data have been shown to rely less on semantically irrelevant words and to generalize better out-of-domain. While this work draws on causal thinking, casting edits as interventions and relying on human understanding to assess outcomes, the underlying causal model is not clear nor are the principles underlying the observed improvements in out-of-domain evaluation. In this paper, we explore a toy analog, using linear Gaussian models. Our analysis reveals interesting relationships between causal models, measurement noise, out-of-domain generalization, and reliance on spurious signals. Interestingly our analysis suggests that data corrupted by adding noise to causal features will degrade out-of-domain performance, while noise added to non-causal features may make models more robust out-of-domain. This analysis yields interesting insights that help to explain the efficacy of counterfactually augmented data. Finally, we present a large-scale empirical study that supports this hypothesis.

Measuring Forecasting Skill from Text

People vary in their ability to make accurate predictions about the future. Prior studies have shown that some individuals can predict the outcome of future events with consistently better accuracy. This leads to a natural question: what makes some forecasters better than others? In this paper we explore connections between the language people use to describe their predictions and their forecasting skill. Datasets from two different forecasting domains are explored: (1) geopolitical forecasts from Good Judgment Open, an online prediction forum and (2) a corpus of company earnings forecasts made by financial analysts. We present a number of linguistic metrics which are computed over text associated with people's predictions about the future including: uncertainty, readability, and emotion. By studying linguistic factors associated with predictions, we are able to shed some light on the approach taken by skilled forecasters. Furthermore, we demonstrate that it is possible to accurately predict forecasting skill using a model that is based solely on language. This could potentially be useful for identifying accurate predictions or potentially skilled forecasters earlier.