• Publications
  • Influence
Convolutional Neural Networks for Sentence Classification
  • Yoon Kim
  • Computer Science
  • EMNLP
  • 25 August 2014
TLDR
We report on a series of experiments with convolutional neural networks (CNN) trained on top of pre-trained word vectors for sentence-level classification tasks. Expand
  • 7,853
  • 1297
  • PDF
Character-Aware Neural Language Models
TLDR
We describe a simple neural language model that relies only on character-level inputs. Expand
  • 1,288
  • 152
  • PDF
Adversarially Regularized Autoencoders
TLDR
We propose a flexible method for training deep latent variable models of discrete structures. Expand
  • 169
  • 49
  • PDF
Temporal Analysis of Language through Neural Language Models
TLDR
We provide a method for automatically detecting change in language across time through a chronologically trained neural language model. Expand
  • 227
  • 32
  • PDF
Sequence-Level Knowledge Distillation
TLDR
We demonstrate that standard knowledge distillation applied to word-level prediction can be effective for NMT, and also introduce two novel sequence-level versions ofknowledge distillation that further improve performance, and somewhat surprisingly, seem to eliminate the need for beam search. Expand
  • 295
  • 27
  • PDF
Structured Attention Networks
TLDR
We show that structured attention networks are simple extensions of the basic attention procedure, and that they allow for extending attention beyond the standard soft-selection approach, such as attending to partial segmentations. Expand
  • 238
  • 22
  • PDF
Semi-Amortized Variational Autoencoders
TLDR
We propose a hybrid approach, to use AVI to initialize the variational parameters and run stochastic variational inference (SVI) to refine them. Expand
  • 131
  • 21
  • PDF
Compound Probabilistic Context-Free Grammars for Grammar Induction
TLDR
We study a formalization of the grammar induction problem that models sentences as being generated by a compound probabilistic context free grammar. Expand
  • 46
  • 19
  • PDF
Unsupervised Recurrent Neural Network Grammars
TLDR
We explore unsupervised learning of recurrent neural network grammars for language modeling and grammar induction. Expand
  • 55
  • 18
  • PDF
Adversarially Regularized Autoencoders for Generating Discrete Structures
TLDR
Generative adversarial networks are an effective approach for learning rich latent representations of continuous data, but have proven difficult to apply directly to discrete structured data, such as discretized images. Expand
  • 74
  • 15
  • PDF
...
1
2
...