• Publications
  • Influence
Globally Normalized Transition-Based Neural Networks
We introduce a globally normalized transition-based neural network model that achieves state-of-the-art part-of-speech tagging, dependency parsing and sentence compression results. Our model is aExpand
Natural Questions: A Benchmark for Question Answering Research
TLDR
The Natural Questions corpus, a question answering data set, is presented, introducing robust metrics for the purposes of evaluating question answering systems; demonstrating high human upper bounds on these metrics; and establishing baseline results using competitive methods drawn from related literature. Expand
BoolQ: Exploring the Surprising Difficulty of Natural Yes/No Questions
TLDR
It is found that transferring from entailment data is more effective than transferring from paraphrase or extractive QA data, and that it, surprisingly, continues to be very beneficial even when starting from massive pre-trained language models such as BERT. Expand
Structured Training for Neural Network Transition-Based Parsing
TLDR
This work presents structured perceptron training for neural network transition-based dependency parsing, and provides indepth ablative analysis to determine which aspects of this model provide the largest gains in accuracy. Expand
TyDi QA: A Benchmark for Information-Seeking Question Answering in Typologically Diverse Languages
TLDR
A quantitative analysis of the data quality and example-level qualitative linguistic analyses of observed language phenomena that would not be found in English-only corpora are presented. Expand
Synthetic QA Corpora Generation with Roundtrip Consistency
TLDR
A novel method of generating synthetic question answering corpora is introduced by combining models of question generation and answer extraction, and by filtering the results to ensure roundtrip consistency, establishing a new state-of-the-art on SQuAD2 and NQ. Expand
A BERT Baseline for the Natural Questions
TLDR
A new baseline for the Natural Questions is described and the gap between the model F1 scores reported in the original dataset paper and the human upper bound is reduced by 30% and 50% relative for the long and short answer tasks respectively. Expand
Fusion of Detected Objects in Text for Visual Question Answering
TLDR
A detailed ablation analysis shows that the early integration of the visual features into the text analysis is key to the effectiveness of the new architecture. Expand
Transforming Dependency Structures to Logical Forms for Semantic Parsing
TLDR
This work introduces a robust system based on the lambda calculus for deriving neo-Davidsonian logical forms from dependency trees and obtains the strongest result to date on Free917 and competitive results on WebQuestions. Expand
Answer Extraction
TLDR
This paper describes a system that attempts to retrieve a much smaller section of text, namely, a direct answer to a user's question, using the SMART IR system to extract a ranked set of passages that are relevant to the query. Expand
...
1
2
...