{"url":"/dataset/wikihop","name":"WikiHop","full_name":null,"description_markdown":"**WikiHop** is a multi-hop question-answering dataset. The query of WikiHop is constructed with entities and relations from WikiData, while supporting documents are from WikiReading. A bipartite graph connecting entities and documents is first built and the answer for each query is located by traversal on this graph. Candidates that are type-consistent with the answer and share the same relation in query with the answer are included, resulting in a set of candidates. Thus, WikiHop is a multi-choice style reading comprehension data set. There are totally about 43K samples in training set, 5K samples in development set and 2.5K samples in test set. The test set is not provided. The task is to predict the correct answer given a query and multiple supporting documents.\r\n\r\nThe dataset includes a masked variant, where all candidates and their mentions in the supporting documents are replaced by random but consistent placeholder tokens.\r\n\r\nSource: [Multi-hop Reading Comprehension across Multiple Documents by Reasoning over Heterogeneous Graphs](https://arxiv.org/abs/1905.07374)\r\nImage Source: [http://qangaroo.cs.ucl.ac.uk/](http://qangaroo.cs.ucl.ac.uk/)","description_withheld":null,"homepage":"http://qangaroo.cs.ucl.ac.uk/","introduced_date":"2017-01-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/constructing-datasets-for-multi-hop-reading","title":"Constructing Datasets for Multi-hop Reading Comprehension Across Documents","first_author":"Johannes Welbl","url":null},"license":{"name":"CC BY-SA 3.0","url":"http://qangaroo.cs.ucl.ac.uk/"},"modalities":[{"name":"Texts","url":"/datasets/modality/texts"}],"tasks":[{"name":"Question Answering","url":"/task/question-answering","datasets_with_task":"/datasets/task/question-answering"},{"name":"Paraphrase Identification","url":"/task/paraphrase-identification","datasets_with_task":"/datasets/task/paraphrase-identification"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["WikiHop"],"data_loaders":[{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/wiki_hop","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/KETI-AIR/wiki_hop","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/KETI-AIR/kor_wiki_hop","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/QAngaroo/wiki_hop","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/allenai/allennlp-models","url":"https://docs.allennlp.org/models/main/models/rc/dataset_readers/qangaroo/","frameworks":["pytorch"]}],"num_papers_in_archive":66,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/question-answering-on-wikihop","task":"Question Answering","dataset_variant":"WikiHop","rows":9,"metrics":["Test"],"first_row_in_archive_order":{"model":"BigBird-etc","paper":"/paper/big-bird-transformers-for-longer-sequences","metrics":{"Test":"82.3"},"code_links":[{"title":"huggingface/transformers","url":"https://github.com/huggingface/transformers"},{"title":"tensorflow/models","url":"https://github.com/tensorflow/models/tree/master/official/nlp/projects/bigbird"},{"title":"PaddlePaddle/PaddleNLP","url":"https://github.com/PaddlePaddle/PaddleNLP/tree/develop/paddlenlp/transformers/bigbird"},{"title":"facebookresearch/xformers","url":"https://github.com/facebookresearch/xformers"},{"title":"google-research/bigbird","url":"https://github.com/google-research/bigbird"},{"title":"monologg/kobigbird","url":"https://github.com/monologg/kobigbird"},{"title":"mim-solutions/bert_for_longer_texts","url":"https://github.com/mim-solutions/bert_for_longer_texts"},{"title":"mim-solutions/roberta_for_longer_texts","url":"https://github.com/mim-solutions/roberta_for_longer_texts"},{"title":"sajjjadayobi/ParsBigBird","url":"https://github.com/sajjjadayobi/ParsBigBird"},{"title":"thefonseca/factorsum","url":"https://github.com/thefonseca/factorsum"},{"title":"2024-MindSpore-1/Code2","url":"https://github.com/2024-MindSpore-1/Code2/tree/main/model-1/big_bird"},{"title":"sergeykramp/mthesis-bigbird-embeddings","url":"https://github.com/sergeykramp/mthesis-bigbird-embeddings"},{"title":"pwc-1/Paper-8","url":"https://github.com/pwc-1/Paper-8/tree/main/big_bird"},{"title":"pwc-1/Paper-8","url":"https://github.com/pwc-1/Paper-8/tree/main/bigbird_pegasus"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/paraphrase-identification-on-wikihop","task":"Paraphrase Identification","dataset_variant":"WikiHop","rows":1,"metrics":["Accuracy"],"first_row_in_archive_order":{"model":"StructBERTRoBERTa ensemble","paper":"/paper/structbert-incorporating-language-structures","metrics":{"Accuracy":"90.7%"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/luke-graph-a-transformer-based-approach-with","title":"LUKE-Graph: A Transformer-based Approach with Gated Relational Graph Attention for Cloze-style Reading Comprehension","date":"2023-03-12","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/big-bird-transformers-for-longer-sequences","title":"Big Bird: Transformers for Longer Sequences","date":"2020-07-28","rows_on_this_dataset":1,"code_links":14,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":15,"samples_ran":10,"samples_unverified":5,"pointer_only_for_licence":11,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/longformer-the-long-document-transformer","title":"Longformer: The Long-Document Transformer","date":"2020-04-10","rows_on_this_dataset":1,"code_links":22,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":35,"samples_ran":15,"samples_unverified":20,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/multi-hop-question-answering-via-reasoning","title":"Multi-hop Question Answering via Reasoning Chains","date":"2019-10-07","rows_on_this_dataset":1,"code_links":3,"syntology":null},{"paper":"/paper/structbert-incorporating-language-structures","title":"StructBERT: Incorporating Language Structures into Pre-training for Deep Language Understanding","date":"2019-08-13","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/coarse-grain-fine-grain-coattention-network","title":"Coarse-grain Fine-grain Coattention Network for Multi-evidence Question Answering","date":"2019-01-03","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/commonsense-for-generative-multi-hop-question","title":"Commonsense for Generative Multi-Hop Question Answering Tasks","date":"2018-09-17","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":7,"samples_ran":0,"samples_unverified":7,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/exploring-graph-structured-passage","title":"Exploring Graph-structured Passage Representation for Multi-hop Reading Comprehension with Graph Neural Networks","date":"2018-09-06","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/neural-models-for-reasoning-over-multiple","title":"Neural Models for Reasoning over Multiple Mentions using Coreference","date":"2018-04-16","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/constructing-datasets-for-multi-hop-reading","title":"Constructing Datasets for Multi-hop Reading Comprehension Across Documents","date":"2017-10-17","rows_on_this_dataset":1,"code_links":0,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":3,"samples_harvested":57,"samples_ran":25,"samples_unverified":32,"pointer_only_for_licence":16,"papers_with_no_sample_that_ran":1,"note":"the per-paper counts above, summed; not a rate"},"papers_note":"The archive never published its papers-using-dataset list; these are papers with a leaderboard row on this dataset's benchmarks."}