{"url":"/dataset/kilt","name":"KILT","full_name":"KILT Benchmark","description_markdown":"**KILT** (**Knowledge Intensive Language Tasks**) is a benchmark consisting of 11 datasets representing 5 types of tasks:\r\n\r\n* Fact-checking (FEVER),\r\n* Entity linking (AIDA CoNLL-YAGO, WNED-WIKI, WNED-CWEB),\r\n* Slot filling (T-Rex, Zero Shot RE),\r\n* Open domain QA (Natural Questions, HotpotQA, TriviaQA, ELI5),\r\n* Dialog generation (Wizard of Wikipedia).\r\n\r\nAll these datasets have been grounded in a single pre-processed wikipedia snapshot, allowing for fairer and more consistent evaluation as well as enabling new task setups such as multitask and transfer learning.\r\n\r\nSource: [KILT Benchmarking](https://ai.facebook.com/tools/kilt/)","description_withheld":null,"homepage":"http://kiltbenchmark.com/","introduced_date":"2020-09-04","introduced_date_note":null,"introduced_by":{"paper":"/paper/kilt-a-benchmark-for-knowledge-intensive","title":"KILT: a Benchmark for Knowledge Intensive Language Tasks","first_author":"Fabio Petroni","url":null},"license":null,"modalities":[{"name":"Texts","url":"/datasets/modality/texts"}],"tasks":[{"name":"Question Answering","url":"/task/question-answering","datasets_with_task":"/datasets/task/question-answering"},{"name":"Entity Linking","url":"/task/entity-linking","datasets_with_task":"/datasets/task/entity-linking"},{"name":"Slot Filling","url":"/task/slot-filling","datasets_with_task":"/datasets/task/slot-filling"},{"name":"Open-Domain Question Answering","url":"/task/open-domain-question-answering","datasets_with_task":"/datasets/task/open-domain-question-answering"},{"name":"Fact Verification","url":"/task/fact-verification","datasets_with_task":"/datasets/task/fact-verification"},{"name":"Open-Domain Dialog","url":"/task/open-domain-dialog","datasets_with_task":"/datasets/task/open-domain-dialog"}],"languages":[],"variants":["NQ KILT","KILT: Zero Shot RE","KILT: Wizard of Wikipedia","KILT: WNED-WIKI","KILT: WNED-CWEB","KILT: TriviaQA","KILT: T-REx","KILT: Natural Questions","KILT: HotpotQA","KILT: FEVER","KILT: ELI5","KILT: AIDA-YAGO2","KILT"],"data_loaders":[{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/facebook/kilt_tasks","frameworks":["tf","pytorch","jax"]},{"repo":"https://github.com/huggingface/datasets","url":"https://huggingface.co/datasets/kilt_tasks","frameworks":["tf","pytorch","jax"]}],"num_papers_in_archive":117,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/fact-verification-on-kilt-fever","task":"Fact Verification","dataset_variant":"KILT: FEVER","rows":33,"metrics":["KILT-AC","R-Prec","Recall@5","Accuracy"],"first_row_in_archive_order":{"model":"Re2G","paper":"/paper/re2g-retrieve-rerank-generate-2","metrics":{"Accuracy":"89.55","KILT-AC":"78.53","R-Prec":"88.92","Recall@5":"92.52"},"code_links":[{"title":"ibm/kgi-slot-filling","url":"https://github.com/ibm/kgi-slot-filling"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/open-domain-dialog-on-kilt-wizard-of","task":"Open-Domain Dialog","dataset_variant":"KILT: Wizard of Wikipedia","rows":21,"metrics":["KILT-RL","R-Prec","Recall@5","ROUGE-L","F1","KILT-F1"],"first_row_in_archive_order":{"model":"Hindsight","paper":null,"metrics":{"F1":"19.19","KILT-F1":"13.39","KILT-RL":"11.92","R-Prec":"56.08","ROUGE-L":"17.06","Recall@5":"74.27"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/slot-filling-on-kilt-zero-shot-re","task":"Slot Filling","dataset_variant":"KILT: Zero Shot RE","rows":21,"metrics":["KILT-AC","R-Prec","Recall@5","Accuracy","F1","KILT-F1"],"first_row_in_archive_order":{"model":"single ngram","paper":null,"metrics":{"Accuracy":"74.63","F1":"79.66","KILT-AC":"73.2","KILT-F1":"78.12","R-Prec":"97.99","Recall@5":"99.34"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/slot-filling-on-kilt-t-rex","task":"Slot Filling","dataset_variant":"KILT: T-REx","rows":20,"metrics":["KILT-AC","R-Prec","Recall@5","Accuracy","F1","KILT-F1"],"first_row_in_archive_order":{"model":"Re2G","paper":"/paper/re2g-retrieve-rerank-generate-2","metrics":{"Accuracy":"87.68","F1":"89.93","KILT-AC":"75.84","KILT-F1":"77.05","R-Prec":"80.7","Recall@5":"89.0"},"code_links":[{"title":"ibm/kgi-slot-filling","url":"https://github.com/ibm/kgi-slot-filling"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/open-domain-question-answering-on-kilt","task":"Open-Domain Question Answering","dataset_variant":"KILT: Natural Questions","rows":16,"metrics":["KILT-EM","R-Prec","Recall@5","EM","F1","KILT-F1"],"first_row_in_archive_order":{"model":"Re2G","paper":"/paper/re2g-retrieve-rerank-generate-2","metrics":{"EM":"51.73","F1":"60.97","KILT-EM":"43.56","KILT-F1":"49.8","R-Prec":"70.78","Recall@5":"76.63"},"code_links":[{"title":"ibm/kgi-slot-filling","url":"https://github.com/ibm/kgi-slot-filling"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/open-domain-question-answering-on-kilt-eli5","task":"Open-Domain Question Answering","dataset_variant":"KILT: ELI5","rows":16,"metrics":["KILT-RL","R-Prec","Recall@5","ROUGE-L","F1","KILT-F1"],"first_row_in_archive_order":{"model":"somebody","paper":null,"metrics":{"F1":"27.13","KILT-F1":"3.0","KILT-RL":"2.62","R-Prec":"10.83","ROUGE-L":"24.53","Recall@5":"27.25"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/open-domain-question-answering-on-kilt-1","task":"Open-Domain Question Answering","dataset_variant":"KILT: HotpotQA","rows":14,"metrics":["KILT-EM","R-Prec","Recall@5","EM","F1","KILT-F1"],"first_row_in_archive_order":{"model":"intersect","paper":null,"metrics":{"EM":"40.46","F1":"51.44","KILT-EM":"18.06","KILT-F1":"21.42","R-Prec":"58.83","Recall@5":"51.03"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/entity-linking-on-kilt-aida-yago2","task":"Entity Linking","dataset_variant":"KILT: AIDA-YAGO2","rows":11,"metrics":["KILT-AC","R-Prec","Recall@5","Accuracy"],"first_row_in_archive_order":{"model":"GENRE","paper":"/paper/autoregressive-entity-retrieval","metrics":{"Accuracy":"89.85","KILT-AC":"89.85","R-Prec":"89.85","Recall@5":"94.76"},"code_links":[{"title":"facebookresearch/GENRE","url":"https://github.com/facebookresearch/GENRE"},{"title":"amzn/seqzero","url":"https://github.com/amzn/seqzero"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/entity-linking-on-kilt-wned-cweb","task":"Entity Linking","dataset_variant":"KILT: WNED-CWEB","rows":10,"metrics":["KILT-AC","R-Prec","Recall@5","Accuracy"],"first_row_in_archive_order":{"model":"GENRE","paper":"/paper/autoregressive-entity-retrieval","metrics":{"Accuracy":"71.22","KILT-AC":"71.22","R-Prec":"71.22","Recall@5":"79.22"},"code_links":[{"title":"facebookresearch/GENRE","url":"https://github.com/facebookresearch/GENRE"},{"title":"amzn/seqzero","url":"https://github.com/amzn/seqzero"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/entity-linking-on-kilt-wned-wiki","task":"Entity Linking","dataset_variant":"KILT: WNED-WIKI","rows":10,"metrics":["KILT-AC","R-Prec","Recall@5","Accuracy"],"first_row_in_archive_order":{"model":"GENRE","paper":"/paper/autoregressive-entity-retrieval","metrics":{"Accuracy":"87.44","KILT-AC":"87.44","R-Prec":"87.44","Recall@5":"94.91"},"code_links":[{"title":"facebookresearch/GENRE","url":"https://github.com/facebookresearch/GENRE"},{"title":"amzn/seqzero","url":"https://github.com/amzn/seqzero"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"},{"leaderboard":"/sota/question-answering-on-kilt-eli5","task":"Question Answering","dataset_variant":"KILT: ELI5","rows":7,"metrics":["Rouge-L","F1"],"first_row_in_archive_order":{"model":"RBG","paper":"/paper/read-before-generate-faithful-long-form","metrics":{"F1":"24.53","Rouge-L":"27.13"},"code_links":[]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/an-efficient-memory-augmented-transformer-for","title":"An Efficient Memory-Augmented Transformer for Knowledge-Intensive NLP Tasks","date":"2022-10-30","rows_on_this_dataset":1,"code_links":1,"syntology":null},{"paper":"/paper/re2g-retrieve-rerank-generate-2","title":"Re2G: Retrieve, Rerank, Generate","date":"2022-07-13","rows_on_this_dataset":4,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":8,"samples_ran":1,"samples_unverified":7,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/knowledge-infused-decoding-1","title":"Knowledge Infused Decoding","date":"2022-04-06","rows_on_this_dataset":1,"code_links":1,"syntology":{"read_at":"2026-09-24T18:15:14+00:00","samples_harvested":3,"samples_ran":0,"samples_unverified":3,"pointer_only_for_licence":0,"claim":"Per-sample execution on synthesized fixtures; not a correctness claim."}},{"paper":"/paper/read-before-generate-faithful-long-form","title":"Read before Generate! Faithful Long Form Question Answering with Machine Reading","date":"2022-03-01","rows_on_this_dataset":1,"code_links":0,"syntology":null},{"paper":"/paper/hurdles-to-progress-in-long-form-question","title":"Hurdles to Progress in Long-form Question Answering","date":"2021-03-10","rows_on_this_dataset":2,"code_links":2,"syntology":null},{"paper":"/paper/learning-dense-representations-of-phrases-at","title":"Learning Dense Representations of Phrases at Scale","date":"2020-12-23","rows_on_this_dataset":2,"code_links":4,"syntology":null},{"paper":"/paper/autoregressive-entity-retrieval","title":"Autoregressive Entity Retrieval","date":"2020-10-02","rows_on_this_dataset":3,"code_links":2,"syntology":null},{"paper":"/paper/kilt-a-benchmark-for-knowledge-intensive","title":"KILT: a Benchmark for Knowledge Intensive Language Tasks","date":"2020-09-04","rows_on_this_dataset":14,"code_links":3,"syntology":{"read_at":"2026-09-24T18:15:14+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."}}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":3,"samples_harvested":12,"samples_ran":2,"samples_unverified":10,"pointer_only_for_licence":1,"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."}