• Publications
  • Influence
Text Understanding with the Attention Sum Reader Network
TLDR
We present a new model that uses attention to directly pick the answer from the context as opposed to computing the answer using a blended representation of words in the document as is usual in similar models. Expand
  • 255
  • 51
  • PDF
Knowledge Base Completion: Baselines Strike Back
TLDR
This paper shows that the accuracy of almost all models published on the FB15k can be outperformed by an appropriately tuned baseline - our reimplementation of the DistMult model. Expand
  • 114
  • 18
  • PDF
Embracing data abundance: BookTest Dataset for Reading Comprehension
TLDR
This article is making a case for the community to move to larger data and as a step in that direction it is proposing a new dataset similar to the popular Children's Book Test (CBT), however more than 60 times larger. Expand
  • 40
  • 8
  • PDF
Planning for Goal-Oriented Dialogue Systems
TLDR
In this work, we propose a paradigm shift in the creation of goal-oriented complex dialogue systems that dramatically eliminates the need for a designer to manually specify a dialogue tree. Expand
  • 6
  • 1
  • PDF
Embracing Data Abundance
TLDR
This article is making a case for the community to move to larger data in Text Comprehension research and is offering the BookTest dataset as a step in that direction. Expand
  • 2
  • 1
Generating Dialogue Agents via Automated Planning
TLDR
We tackle this challenging problem by using domain-independent AI planning to automatically create dialogue plans to guide a dialogue towards achieving a given goal. Expand
  • 6
  • PDF
From Particular to General: A Preliminary Case Study of Transfer Learning in Reading Comprehension
In this paper we argue that transfer learning will be an important ingredient of general learning AI. We are especially interested in using data-rich domains to learn skills widely applicable inExpand
  • 5
  • PDF
Finding a Jack-of-All-Trades: An Examination of Semi-supervised Learning in Reading Comprehension
TLDR
We train a neural-network-based model on two context-question-answer datasets, the BookTest and CNN/Daily Mail, and we monitor transfer to subsets of bAbI, a set of artificial tasks designed to test specific reasoning abilities, and SQuAD, a question-answering dataset which is much closer to real-world applications. Expand
  • 1
A Boo(n) for Evaluating Architecture Performance
TLDR
We point out important problems with the common practice of using the best single model performance for comparing deep learning architectures, and we propose a method that corrects these flaws. Expand
  • 2
  • PDF