We create a new NLI test set that shows the deficiency of state-of-the-art models in inferences that require lexical and world knowledge. The new examples are simpler than the SNLI test set, containing sentences that differ by at most one word from sentences in the training set. Yet, the performance on the new test set is substantially worse across systems trained on SNLI, demonstrating that these systems are limited in their generalization ability, failing to capture many simple inferences. IntroductionRecognizing textual entailment (RTE) (Dagan et al., 2013), recently framed as natural language inference (NLI) (Bowman et al., 2015) is a task concerned with identifying whether a premise sentence entails, contradicts or is neutral with the hypothesis sentence. Following the release of the large-scale SNLI dataset (Bowman et al., 2015), many end-to-end neural models have been developed for the task, achieving high accuracy on the test set. As opposed to previous-generation methods, which relied heavily on lexical resources, neural models only make use of pre-trained word embeddings. The few efforts to incorporate external lexical knowledge resulted in negligible performance gain (Chen et al., 2018). This raises the question whether (1) neural methods are inherently stronger, obviating the need of external lexical knowledge;(2) large-scale training data allows for implicit learning of previously explicit lexical knowledge; or (3) the NLI datasets are simpler than early RTE datasets, requiring less knowledge. 1 The contradiction example follows the assumption in Bowman et al. (2015) that the premise contains the most prominent information in the event, hence the premise can't describe the event of a man holding both instruments.

Detecting hypernymy relations is a key task in NLP, which is addressed in the literature using two complementary approaches. Distributional methods, whose supervised variants are the current best performers, and path-based methods, which received less research attention. We suggest an improved path-based algorithm, in which the dependency paths are encoded using a recurrent neural network, that achieves results comparable to distributional methods. We then extend the approach to integrate both pathbased and distributional signals, significantly improving upon the state-of-the-art on this task.

Recognizing coreferring events and entities across multiple texts is crucial for many NLP applications. Despite the task's importance, research focus was given mostly to withindocument entity coreference, with rather little attention to the other variants. We propose a neural architecture for cross-document coreference resolution. Inspired by Lee et al.(2012), we jointly model entity and event coreference. We represent an event (entity) mention using its lexical span, surrounding context, and relation to entity (event) mentions via predicate-arguments structures. Our model outperforms the previous state-of-the-art event coreference model on ECB+, while providing the first entity coreference results on this corpus. Our analysis confirms that all our representation elements, including the mention span itself, its context, and the relation to other mentions contribute to the model's success.

Natural language understanding involves reading between the lines with implicit background knowledge. Current systems either rely on pretrained language models as the sole implicit source of world knowledge, or resort to external knowledge bases (KBs) to incorporate additional relevant knowledge. We propose an unsupervised framework based on self-talk as a novel alternative to multiple-choice commonsense tasks. Inspired by inquiry-based discovery learning (Bruner, 1961), our approach inquires language models with a number of information seeking questions such as "what is the definition of ..." to discover additional background knowledge. Empirical results demonstrate that the self-talk procedure substantially improves the performance of zeroshot language model baselines on four out of six commonsense benchmarks, and competes with models that obtain knowledge from external KBs. While our approach improves performance on several benchmarks, the selftalk induced knowledge even when leading to correct answers is not always seen as helpful by human judges, raising interesting questions about the inner-workings of pre-trained language models for commonsense reasoning.

The fundamental role of hypernymy in NLP has motivated the development of many methods for the automatic identification of this relation, most of which rely on word distribution. We investigate an extensive number of such unsupervised measures, using several distributional semantic models that differ by context type and feature weighting. We analyze the performance of the different methods based on their linguistic motivation. Comparison to the state-of-the-art supervised methods shows that while supervised methods generally outperform the unsupervised ones, the former are sensitive to the distribution of training instances, hurting their reliability. Being based on general linguistic hypotheses and independent from training data, unsupervised measures are more robust, and therefore are still useful artillery for hypernymy detection.

Building meaningful phrase representations is challenging because phrase meanings are not simply the sum of their constituent meanings. Lexical composition can shift the meanings of the constituent words and introduce implicit information. We tested a broad range of textual representations for their capacity to address these issues. We found that, as expected, contextualized word representations perform better than static word embeddings, more so on detecting meaning shift than in recovering implicit information, in which their performance is still far from that of humans. Our evaluation suite, consisting of six tasks related to lexical composition effects, can serve future research aiming to improve representations.

Abductive and counterfactual reasoning, core abilities of everyday human cognition, require reasoning about what might have happened at time t, while conditioning on multiple contexts from the relative past and future. However, simultaneous incorporation of past and future contexts using generative language models (LMs) can be challenging, as they are trained either to condition only on the past context or to perform narrowly scoped text-infilling.In this paper, we propose DELOREAN, a new unsupervised decoding algorithm that can flexibly incorporate both the past and future contexts using only off-the-shelf, left-to-right language models and no supervision. The key intuition of our algorithm is incorporating the future through back-propagation, during which, we only update the internal representation of the output while fixing the model parameters. By alternating between forward and backward propagation, DELOREAN can decode the output representation that reflects both the left and right contexts. We demonstrate that our approach is general and applicable to two nonmonotonic reasoning tasks: abductive text generation and counterfactual story revision, where DELOREAN outperforms a range of unsupervised and some supervised methods, based on automatic and human evaluation. 1

Defeasible inference is a mode of reasoning in which an inference (X is a bird, therefore X flies) may be weakened or overturned in light of new evidence (X is a penguin). Though long recognized in classical AI and philosophy, defeasible inference has not been extensively studied in the context of contemporary data-driven research on natural language inference and commonsense reasoning. We introduce Defeasible NLI (abbreviated δ-NLI), a dataset for defeasible inference in natural language. δ-NLI contains extensions to three existing inference datasets covering diverse modes of reasoning: common sense, natural language inference, and social norms. From δ-NLI, we develop both a classification and generation task for defeasible inference, and demonstrate that the generation task is much more challenging. Despite lagging human performance, however, generative models trained on this data are capable of writing sentences that weaken or strengthen a specified inference up to 68% of the time.