Measuring what linguistic information is encoded in neural models of language has become popular in NLP. Researchers approach this enterprise by training "probes" - supervised models designed to extract linguistic structure from another model's output. One such probe is the structural probe (Hewitt and Manning, 2019), designed to quantify the extent to which syntactic information is encoded in contextualised word representations. The structural probe has a novel design, unattested in the parsing literature, the precise benefit of which is not immediately obvious. To explore whether syntactic probes would do better to make use of existing techniques, we compare the structural probe to a more traditional parser with an identical lightweight parameterisation. The parser outperforms structural probe on UUAS in seven of nine analysed languages, often by a substantial amount (e.g. by 11.1 points in English). Under a second less common metric, however, there is the opposite trend - the structural probe outperforms the parser. This begs the question: which metric should we prefer?
We use large-scale corpora in six different gendered languages, along with tools from NLP and information theory, to test whether there is a relationship between the grammatical genders of inanimate nouns and the adjectives used to describe those nouns. For all six languages, we find that there is a statistically significant relationship. We also find that there are statistically significant relationships between the grammatical genders of inanimate nouns and the verbs that take those nouns as direct objects, as indirect objects, and as subjects. We defer a deeper investigation of these relationships for future work.
The noun lexica of many natural languages are divided into several declension classes with characteristic morphological properties. Class membership is far from deterministic, but the phonological form of a noun and/or its meaning can often provide imperfect clues. Here, we investigate the strength of those clues. More specifically, we operationalize this by measuring how much information, in bits, we can glean about declension class from knowing the form and/or meaning of nouns. We know that form and meaning are often also indicative of grammatical gender---which, as we quantitatively verify, can itself share information with declension class---so we also control for gender. We find for two Indo-European languages (Czech and German) that form and meaning respectively share significant amounts of information with class (and contribute additional information above and beyond gender). The three-way interaction between class, form, and meaning (given gender) is also significant. Our study is important for two reasons: First, we introduce a new method that provides additional quantitative support for a classic linguistic finding that form and meaning are relevant for the classification of nouns into declensions. Secondly, we show not only that individual declensions classes vary in the strength of their clues within a language, but also that these variations themselves vary across languages. The code is publicly available at https://github.com/rycolab/declension-mi.
Machine learning models are trained to find patterns in data. NLP models can inadvertently learn socially undesirable patterns when training on gender biased text. In this work, we propose a general framework that decomposes gender bias in text along several pragmatic and semantic dimensions: bias from the gender of the person being spoken about, bias from the gender of the person being spoken to, and bias from the gender of the speaker. Using this fine-grained framework, we automatically annotate eight large scale datasets with gender information. In addition, we collect a novel, crowdsourced evaluation benchmark of utterance-level gender rewrites. Distinguishing between gender bias along multiple dimensions is important, as it enables us to train finer-grained gender bias classifiers. We show our classifiers prove valuable for a variety of important applications, such as controlling for gender bias in generative models, detecting gender bias in arbitrary text, and shed light on offensive language in terms of genderedness.
Natural language inference (NLI) is an increasingly important task for natural language understanding, which requires one to infer whether one sentence entails another. However, the ability of NLI models to make pragmatic inferences remains understudied. We create an IMPlicature and PRESupposition diagnostic dataset (IMPPRES), consisting of 32K semi-automatically generated sentence pairs illustrating well-studied pragmatic inference types. We use IMPPRES to evaluate whether BERT, BOW, and InferSent NLI models trained on MultiNLI (Williams et al., 2018) learn to make pragmatic inferences. Although MultiNLI contains vanishingly few pairs illustrating these inference types, we find that BERT learns to draw pragmatic inferences: it reliably treats implicatures triggered by "some" as entailments. For some presupposition triggers like "only", BERT reliably recognizes the presupposition as an entailment, even when the trigger is embedded under an entailment canceling operator like negation. BOW and InferSent show weaker evidence of pragmatic reasoning. We conclude that NLI training encourages models to learn some, but not all, pragmatic inferences.
The success of neural networks on a diverse set of NLP tasks has led researchers to question how much do these networks actually know about natural language. Probes are a natural way of assessing this. When probing, a researcher chooses a linguistic task and trains a supervised model to predict annotation in that linguistic task from the network's learned representations. If the probe does well, the researcher may conclude that the representations encode knowledge related to the task. A commonly held belief is that using simpler models as probes is better; the logic is that such models will identify linguistic structure, but not learn the task itself. We propose an information-theoretic formalization of probing as estimating mutual information that contradicts this received wisdom: one should always select the highest performing probe one can, even if it is more complex, since it will result in a tighter estimate. The empirical portion of our paper focuses on obtaining tight estimates for how much information BERT knows about parts of speech in a set of five typologically diverse languages that are often underrepresented in parsing research, plus English, totaling six languages. We find BERT accounts for only at most 5% more information than traditional, type-based word embeddings.
Models often easily learn biases present in the training data, and their predictions directly reflect this bias. We analyze the presence of gender bias in dialogue and examine the subsequent effect on generative chitchat dialogue models. Based on this analysis, we propose a combination of three techniques to mitigate bias: counterfactual data augmentation, targeted data collection, and conditional training. We focus on the multi-player text-based fantasy adventure dataset LIGHT as a testbed for our work. LIGHT contains gender imbalance between male and female characters with around 1.6 times as many male characters, likely because it is entirely collected by crowdworkers and reflects common biases that exist in fantasy or medieval settings. We show that (i) our proposed techniques mitigate gender bias by balancing the genderedness of generated dialogue utterances; and (ii) they work particularly well in combination. Further, we show through various metrics---such as quantity of gendered words, a dialogue safety classifier, and human evaluation---that our models generate less gendered, but still engaging chitchat responses.
We introduce a new large-scale NLI benchmark dataset, collected via an iterative, adversarial human-and-model-in-the-loop procedure. We show that training models on this new dataset leads to state-of-the-art performance on a variety of popular NLI benchmarks, while posing a more difficult challenge with its new test set. Our analysis sheds light on the shortcomings of current state-of-the-art models, and shows that non-expert annotators are successful at finding their weaknesses. The data collection method can be applied in a never-ending learning scenario, becoming a moving target for NLU, rather than a static benchmark that will quickly saturate.
Many of the world's languages employ grammatical gender on the lexeme. For example, in Spanish, the word for 'house' (casa) is feminine, whereas the word for 'paper' (papel) is masculine. To a speaker of a genderless language, this assignment seems to exist with neither rhyme nor reason. But is the assignment of inanimate nouns to grammatical genders truly arbitrary? We present the first large-scale investigation of the arbitrariness of noun-gender assignments. To that end, we use canonical correlation analysis to correlate the grammatical gender of inanimate nouns with an externally grounded definition of their lexical semantics. We find that 18 languages exhibit a significant correlation between grammatical gender and lexical semantics.
While idiosyncrasies of the Chinese classifier system have been a richly studied topic among linguists (Adams and Conklin, 1973; Erbaugh, 1986; Lakoff, 1986), not much work has been done to quantify them with statistical methods. In this paper, we introduce an information-theoretic approach to measuring idiosyncrasy; we examine how much the uncertainty in Mandarin Chinese classifiers can be reduced by knowing semantic information about the nouns that the classifiers modify. Using the empirical distribution of classifiers from the parsed Chinese Gigaword corpus (Graff et al., 2005), we compute the mutual information (in bits) between the distribution over classifiers and distributions over other linguistic quantities. We investigate whether semantic classes of nouns and adjectives differ in how much they reduce uncertainty in classifier choice, and find that it is not fully idiosyncratic; while there are no obvious trends for the majority of semantic classes, shape nouns reduce uncertainty in classifier choice the most.