Natural Questions and TriviaQA

The dataset used in the paper is Natural Questions and TriviaQA, two popular datasets for open-domain question answering.

BibTex: