536 datasets found

WebSRC dataset for web-based structural reading comprehension.

Topological Information Enhanced model (TIE) for web-based structural reading comprehension on web pages.

Educational Question Generation (QG) dataset, used to train and evaluate QG models.

Long document question answering (DocQA) aims to answer questions from long documents over 10k words. They usually contain content structures such as sections, sub-sections, and...

The dataset for Difficulty-controllable Question Generation (DQG) for reading comprehension, prepared by the authors.

This dataset is used in the paper to evaluate the performance of the TranSTR architecture.

AR-LSAT: A dataset that collects all analytical logic reasoning questions from the Law School Admission Test

ProofWriter: Generating implications, proofs, and abductive statements over natural language

The dataset used in the paper for natural language reasoning tasks

Video question answering (VideoQA) is an essential task in vision-language understanding, which has attracted numerous research attention recently. Nevertheless, existing works...

The dataset used in this paper is a set of categorical probability distributions for a finite set of categories A = {a1,..., ak}. The dataset is used to evaluate the proposed...

A face anti-spoofing dataset with rich annotations, focusing on questions with a single entity and relation.

The ENsEN dataset is used to evaluate the usefulness of semantic snippets. It contains 10 tasks, each with three questions on a common topic.

The dataset is used to evaluate the query-biased ranking of LOD resources. It contains 30 queries, 150 HTML Web pages, and 81 detected resources by Web page.

The dataset used in the paper is a collection of sentences for the task of pronoun resolution.

The dataset used in the paper is a collection of sentences for the tasks of subject-verb agreement and anaphora resolution.

The ProKnow-data dataset is a collection of diagnostic conversations guided by safety constraints and ProKnow that healthcare professionals use.

The CRASS dataset is a novel basis to test reasoning and natural language understanding of large language models. It consists of 274 high-quality premise-counterfactual tuples,...

The STAR dataset provides 60K situated reasoning questions based on 22K trimmed situation video clips.

The AGQA benchmark is a visual dataset comprising 192M hand-crafted questions about 9.6K videos from the Charades dataset.