{"url":"/dataset/hotpotqa","name":"HotpotQA","full_name":null,"description_markdown":"**HotpotQA** is a question answering dataset collected on the English Wikipedia, containing about 113K crowd-sourced questions that are constructed to require the introduction paragraphs of two Wikipedia articles to answer. Each question in the dataset comes with the two gold paragraphs, as well as a list of sentences in these paragraphs that crowdworkers identify as supporting facts necessary to answer the question. \r\n\r\nA diverse range of reasoning strategies are featured in HotpotQA, including questions involving missing entities in the question, intersection questions (What satisfies property A and property B?), and comparison questions, where two entities are compared by a common attribute, among others. In the few-document distractor setting, the QA models are given ten paragraphs in which the gold paragraphs are guaranteed to be found; in the open-domain fullwiki setting, the models are only given the question and the entire Wikipedia. Models are evaluated on their answer accuracy and explainability, where the former is measured as overlap between the predicted and gold answers with exact match (EM) and unigram F1, and the latter concerns how well the predicted supporting fact sentences match human annotation (Supporting Fact EM/F1). A joint metric is also reported on this dataset, which encourages systems to perform well on both tasks simultaneously.\r\n\r\nSource: [Answering Complex Open-domain Questions Through Iterative Query Generation](https://arxiv.org/abs/1910.07000)\r\nImage Source: [Yang et al](https://arxiv.org/pdf/1809.09600v1.pdf)","description_withheld":null,"homepage":"https://hotpotqa.github.io/","introduced_date":"2018-01-01","introduced_date_note":null,"introduced_by":{"paper":"/paper/hotpotqa-a-dataset-for-diverse-explainable","title":"HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering","first_author":"Zhilin Yang","url":null},"license":{"name":"CC BY-SA 4.0","url":"https://creativecommons.org/licenses/by-sa/4.0/"},"modalities":[{"name":"Texts","url":"/datasets/modality/texts"}],"tasks":[{"name":"Question Answering","url":"/task/question-answering","datasets_with_task":"/datasets/task/question-answering"},{"name":"Reading Comprehension","url":"/task/reading-comprehension","datasets_with_task":"/datasets/task/reading-comprehension"},{"name":"Retrieval","url":"/task/retrieval","datasets_with_task":"/datasets/task/retrieval"},{"name":"Text Retrieval","url":"/task/text-retrieval","datasets_with_task":"/datasets/task/text-retrieval"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["HotpotQA","hotpot_qa"],"data_loaders":[{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/vincentkoc/hotpot_qa_archive","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/hotpotqa/hotpot_qa","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/hotpot_qa","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/facebookresearch/ParlAI","url":"https://parl.ai/docs/tasks.html#hotpotqa","frameworks":["pytorch"]}],"num_papers_in_archive":933,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/question-answering-on-hotpotqa","task":"Question Answering","dataset_variant":"HotpotQA","rows":72,"metrics":["JOINT-F1","ANS-EM","ANS-F1","SUP-EM","SUP-F1","JOINT-EM"],"first_row_in_archive_order":{"model":"Beam Retrieval","paper":"/paper/beam-retrieval-general-end-to-end-retrieval","metrics":{"ANS-EM":"0.727","ANS-F1":"0.850","JOINT-EM":"0.505","JOINT-F1":"0.775","SUP-EM":"0.663","SUP-F1":"0.901"},"code_links":[{"title":"ShayekhBinIslam/openrag","url":"https://github.com/ShayekhBinIslam/openrag"},{"title":"canghongjian/beam_retriever","url":"https://github.com/canghongjian/beam_retriever"},{"title":"Alab-NII/2wikimultihop","url":"https://github.com/Alab-NII/2wikimultihop"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/retrieval-on-hotpotqa","task":"Retrieval","dataset_variant":"HotpotQA","rows":3,"metrics":["Queries per second"],"first_row_in_archive_order":{"model":"BM25S","paper":"/paper/bm25s-orders-of-magnitude-faster-lexical","metrics":{"Queries per second":"20.88"},"code_links":[{"title":"xhluca/bm25s","url":"https://github.com/xhluca/bm25s"},{"title":"xhluca/bm25-benchmarks","url":"https://github.com/xhluca/bm25-benchmarks"},{"title":"conda-forge/bm25s-feedstock","url":"https://github.com/conda-forge/bm25s-feedstock"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/text-retrieval-on-hotpotqa","task":"Text Retrieval","dataset_variant":"HotpotQA","rows":1,"metrics":["nDCG@10"],"first_row_in_archive_order":{"model":"Lucene (BM25S)","paper":"/paper/bm25s-orders-of-magnitude-faster-lexical","metrics":{"nDCG@10":"62.9"},"code_links":[{"title":"xhluca/bm25s","url":"https://github.com/xhluca/bm25s"},{"title":"xhluca/bm25-benchmarks","url":"https://github.com/xhluca/bm25-benchmarks"},{"title":"conda-forge/bm25s-feedstock","url":"https://github.com/conda-forge/bm25s-feedstock"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/bm25s-orders-of-magnitude-faster-lexical","title":"BM25S: Orders of magnitude faster lexical search via eager sparse scoring","date":"2024-07-04","rows_on_this_dataset":4,"code_links":3,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":22,"samples_ran":15,"samples_unverified":7,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/beam-retrieval-general-end-to-end-retrieval","title":"End-to-End Beam Retrieval for Multi-Hop Question Answering","date":"2023-08-17","rows_on_this_dataset":1,"code_links":3,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":13,"samples_ran":10,"samples_unverified":3,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/chain-of-skills-a-configurable-model-for-open","title":"Chain-of-Skills: A Configurable Model for Open-domain Question Answering","date":"2023-05-04","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/adaptive-information-seeking-for-open-domain","title":"Adaptive Information Seeking for Open-Domain Question Answering","date":"2021-09-14","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/hopretriever-retrieve-hops-over-wikipedia-to","title":"HopRetriever: Retrieve Hops over Wikipedia to Answer Complex Questions","date":"2020-12-31","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/retrieve-rerank-read-then-iterate-answering","title":"Answering Open-Domain Questions of Varying Reasoning Steps from Text","date":"2020-10-23","rows_on_this_dataset":2,"code_links":1,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":3,"samples_ran":3,"samples_unverified":0,"pointer_only_for_licence":3,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/answering-complex-open-domain-questions-with","title":"Answering Complex Open-Domain Questions with Multi-Hop Dense Retrieval","date":"2020-09-27","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/ddrqa-dynamic-document-reranking-for-open","title":"Answering Any-hop Open-domain Questions with Iterative Document Reranking","date":"2020-09-16","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":"2026-09-25T09:33:49+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/transformer-xh-multi-evidence-reasoning-with","title":"Transformer-XH: Multi-Evidence Reasoning with eXtra Hop Attention","date":"2020-05-01","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/a-simple-yet-strong-pipeline-for-hotpotqa","title":"A Simple Yet Strong Pipeline for HotpotQA","date":"2020-04-14","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/learning-to-retrieve-reasoning-paths-over-1","title":"Learning to Retrieve Reasoning Paths over Wikipedia Graph for Question Answering","date":"2019-11-24","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":6,"samples_ran":3,"samples_unverified":3,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/hierarchical-graph-network-for-multi-hop","title":"Hierarchical Graph Network for Multi-hop Question Answering","date":"2019-11-09","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":6,"samples_ran":3,"samples_unverified":3,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/multi-paragraph-reasoning-with-knowledge","title":"Multi-Paragraph Reasoning with Knowledge-enhanced Graph Neural Network","date":"2019-11-06","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/answering-complex-open-domain-questions","title":"Answering Complex Open-domain Questions Through Iterative Query Generation","date":"2019-10-15","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/revealing-the-importance-of-semantic","title":"Revealing the Importance of Semantic Retrieval for Machine Reading at Scale","date":"2019-09-17","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":17,"samples_ran":13,"samples_unverified":4,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/multi-hop-paragraph-retrieval-for-open-domain","title":"Multi-Hop Paragraph Retrieval for Open-Domain Question Answering","date":"2019-06-15","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":5,"samples_ran":0,"samples_unverified":5,"pointer_only_for_licence":5,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/multi-hop-reading-comprehension-through","title":"Multi-hop Reading Comprehension through Question Decomposition and Rescoring","date":"2019-06-07","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":1,"samples_ran":1,"samples_unverified":0,"pointer_only_for_licence":1,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/answering-while-summarizing-multi-task","title":"Answering while Summarizing: Multi-task Learning for Multi-hop QA with Evidence Extraction","date":"2019-05-21","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/dynamically-fused-graph-network-for-multi-hop","title":"Dynamically Fused Graph Network for Multi-hop Reasoning","date":"2019-05-16","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":3,"samples_ran":1,"samples_unverified":2,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/cognitive-graph-for-multi-hop-reading","title":"Cognitive Graph for Multi-Hop Reading Comprehension at Scale","date":"2019-05-14","rows_on_this_dataset":1,"code_links":2,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":9,"samples_ran":5,"samples_unverified":4,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/hotpotqa-a-dataset-for-diverse-explainable","title":"HotpotQA: A Dataset for Diverse, Explainable Multi-hop Question Answering","date":"2018-09-25","rows_on_this_dataset":2,"code_links":2,"syntology":{"read_at":"2026-09-25T09:33:49+00:00","samples_harvested":8,"samples_ran":3,"samples_unverified":5,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}}],"syntology_totals":{"read_at":"2026-09-25T09:33:49+00:00","papers_with_samples":13,"samples_harvested":109,"samples_ran":68,"samples_unverified":41,"pointer_only_for_licence":26,"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."}