Dataset Card for PopQA
Dataset Summary
PopQA is a large-scale open-domain question answering (QA) dataset, consisting of 14k entity-centric QA pairs. Each question is created by converting a knowledge tuple retrieved from Wikidata using a template. Each question come with the original subject_entitiey
, object_entity
and relationship_type
annotation, as well as Wikipedia monthly page views.
Languages
The dataset contains samples in English only.
Dataset Structure
Data Instances
- Size of downloaded dataset file: 5.2 MB
Data Fields
id
: question idsubj
: subject entity nameprop
: relationship typeobj
: object entity namesubj_id
: Wikidata ID of the subject entityprop_id
: Wikidata relationship type IDobj_id
: Wikidata ID of the object entitys_aliases
: aliases of the subject entityo_aliases
: aliases of the object entitys_uri
: Wikidata URI of the subject entityo_uri
: Wikidata URI of the object entitys_wiki_title
: Wikipedia page title of the subject entityo_wiki_title
: Wikipedia page title of the object entitys_pop
: Wikipedia monthly pageview of the subject entityo_pop
: Wikipedia monthly pageview of the object entityquestion
: PopQA questionpossible_answers
: a list of the gold answers.
Citation Information
@article{ mallen2023llm_memorization ,
title={When Not to Trust Language Models: Investigating Effectiveness and Limitations of Parametric and Non-Parametric Memories },
author={ Mallen, Alex and Asai,Akari and Zhong, Victor and Das, Rajarshi and Hajishirzi, Hannaneh and Khashabi, Daniel},
journal={ arXiv preprint },
year={ 2022 }
}