Skip to main content ->
Ai2

Research - Papers

Explore a selection of our published work on a variety of key research challenges in AI.

Filter papers

DiscoFuse: A Large-Scale Dataset for Discourse-based Sentence Fusion

Mor GevaEric MalmiIdan SzpektorJonathan Berant
2019
NAACL

Sentence fusion is the task of joining several independent sentences into a single coherent text. Current datasets for sentence fusion are small and insufficient for training modern neural models.… 

Evaluating Text GANs as Language Models

Guy TevetGavriel HabibVered ShwartzJonathan Berant
2019
NAACL

Generative Adversarial Networks (GANs) are a promising approach for text generation that, unlike traditional language models (LM), does not suffer from the problem of “exposure bias”. However, A… 

Linguistic Knowledge and Transferability of Contextual Representations

Nelson F. LiuMatt GardnerYonatan BelinkovNoah A. Smith
2019
NAACL

Contextual word representations derived from large-scale neural language models are successful across a diverse set of NLP tasks, suggesting that they encode useful and transferable features of… 

Polyglot Contextual Representations Improve Crosslingual Transfer

Phoebe MulcaireJungo KasaiNoah A. Smith
2019
NAACL

We introduce a method to produce multilingual contextual word representations by training a single language model on text from multiple languages. Our method combines the advantages of contextual… 

Step-by-Step: Separating Planning from Realization in Neural Data-to-Text Generation

Amit Mor-YosefIdo DaganYoav Goldberg
2019
NAACL

Data-to-text generation can be conceptually divided into two parts: ordering and structuring the information (planning), and generating fluent language describing the information (realization).… 

Aligning Vector-spaces with Noisy Supervised Lexicons

Noa YehezkelJacob GoldbergerYoav Goldberg
2019
NAACL

The problem of learning to translate between two vector spaces given a set of aligned points arises in several application areas of NLP. Current solutions assume that the lexicon which defines the… 

White-to-Black: Efficient Distillation of Black-Box Adversarial Attacks

Or GorodisskyYoav ChaiYotam GilJonathan Berant
2019
NAACL

We show that a neural network can learn to imitate the optimization process performed by white-box attack in a much more efficient manner. We train a black-box attack through this imitation process… 

Repurposing Entailment for Multi-Hop Question Answering Tasks

Harsh TrivediHeeyoung KwonTushar KhotNiranjan Balasubramanian
2019
NAACL

Question Answering (QA) naturally reduces to an entailment problem, namely, verifying whether some text entails the answer to a question. However, for multi-hop QA tasks, which require reasoning… 

Structural Scaffolds for Citation Intent Classification in Scientific Publications

Arman CohanWaleed AmmarMadeleine van ZuylenField Cady
2019
NAACL

Identifying the intent of a citation in scientific papers (e.g., background information, use of methods, comparing results) is critical for machine reading of individual publications and automated… 

Combining Distant and Direct Supervision for Neural Relation Extraction

Iz BeltagyKyle LoWaleed Ammar
2019
NAACL

In relation extraction with distant supervision, noisy labels make it difficult to train quality models. Previous neural models addressed this problem using an attention mechanism that attends to…