We introduce three memory-augmented Recurrent Neural Networks (MARNNs) and explore their capabilities on a series of simple language modeling tasks whose solutions require stack-based mechanisms.
Popular Natural Language Inference (NLI) datasets have been shown to be tainted by hypothesis-only biases.
In contrast to standard approaches to NLI, our methods predict the probability of a premise given a hypothesis and NLI label, discouraging models from ignoring the premise.
In this paper, we systematically assess the ability of standard recurrent networks to perform dynamic counting and to encode hierarchical representations.
While neural, encoder-decoder models have had significant empirical success in text generation, there remain several unaddressed problems with this style of generation.
In a controlled experiment of sequence-to-sequence approaches for the task of sentence correction, we find that character-based models are generally more effective than word-based models and models that encode subword information via convolutions, and that modeling the output data as a series of diffs improves effectiveness over standard approaches.
Recent neural models have shown significant progress on the problem of generating short descriptive texts conditioned on a small number of database records.
We demonstrate that an attention-based encoder-decoder model can be used for sentence-level grammatical error identification for the Automated Evaluation of Scientific Writing (AESW) Shared Task 2016.
There is compelling evidence that coreference prediction would benefit from modeling global information about entity-clusters.
Ranked #15 on Coreference Resolution on OntoNotes