Textual entailment models are increasingly applied in settings like fact-checking, presupposition verification in question answering, or summary evaluation. However, these represent a significant domain shift from existing entailment datasets, and models underperform as a result. We propose WiCE, a new fine-grained textual entailment dataset built on natural claim and evidence pairs extracted from Wikipedia. In addition to standard claim-level entailment, WiCE provides entailment judgments over sub-sentence units of the claim, and a minimal subset of evidence sentences that support each subclaim. To support this, we propose an automatic claim decomposition strategy using GPT-3.5 which we show is also effective at improving entailment models’ performance on multiple datasets at test time. Finally, we show that real claims in our dataset involve challenging verification and retrieval problems that existing models fail to address.

Machine-generated text presents a potential threat not only to the public sphere, but also to the scientific enterprise, whereby genuine research is undermined by convincing, synthetic text. In this paper we examine the problem of detecting GPT-2-generated technical research text. We first consider the realistic scenario where the defender does not have full information about the adversary’s text generation pipeline, but is able to label small amounts of in-domain genuine and synthetic text in order to adapt to the target distribution. Even in the extreme scenario of adapting a physics-domain detector to a biomedical detector, we find that only a few hundred labels are sufficient for good performance. Finally, we show that paragraph-level detectors can be used to detect the tampering of full-length documents under a variety of threat models.

We introduce GEM, a living benchmark for natural language Generation (NLG), its Evaluation, and Metrics. Measuring progress in NLG relies on a constantly evolving ecosystem of automated metrics, datasets, and human evaluation standards. Due to this moving target, new models often still evaluate on divergent anglo-centric corpora with well-established, but flawed, metrics. This disconnect makes it challenging to identify the limitations of current models and opportunities for progress. Addressing this limitation, GEM provides an environment in which models can easily be applied to a wide set of tasks and in which evaluation strategies can be tested. Regular updates to the benchmark will help NLG research become more multilingual and evolve the challenge alongside models. This paper serves as the description of the data for the 2021 shared task at the associated GEM Workshop.

Developing documentation guidelines and easy-to-use templates for datasets and models is a challenging task, especially given the variety of backgrounds, skills, and incentives of the people involved in the building of natural language processing (NLP) tools. Nevertheless, the adoption of standard documentation practices across the field of NLP promotes more accessible and detailed descriptions of NLP datasets and models, while supporting researchers and developers in reflecting on their work. To help with the standardization of documentation, we present two case studies of efforts that aim to develop reusable documentation templates – the HuggingFace data card, a general purpose card for datasets in NLP, and the GEM benchmark data and model cards with a focus on natural language generation. We describe our process for developing these templates, including the identification of relevant stakeholder groups, the definition of a set of guiding principles, the use of existing templates as our foundation, and iterative revisions based on feedback.

We formulate the problem of hypernym prediction as a sequence generation task, where the sequences are taxonomy paths in WordNet. Our experiments with encoder-decoder models show that training to generate taxonomy paths can improve the performance of direct hypernym prediction. As a simple but powerful model, the hypo2path model achieves state-of-the-art performance, outperforming the best benchmark by 4.11 points in hit-at-one (H@1).

In this reproduction paper, we replicate and extend several past studies on transfer learning for entity recognition. In particular, we are interested in entity recognition problems where the class labels in the source and target domains are different. Our work is the first direct comparison of these previously published approaches in this problem setting. In addition, we perform experiments on seven new source/target corpus pairs, nearly doubling the total number of corpus pairs that have been studied in all past work combined. Our results empirically demonstrate when each of the published approaches tends to do well. In particular, simpler approaches often work best when there is very little labeled target data, while neural transfer approaches tend to do better when there is more labeled target data.