{"url":"/dataset/prosocialdialog","name":"ProsocialDialog","full_name":null,"description_markdown":"Most existing dialogue systems fail to respond properly to potentially unsafe user utterances by either ignoring or passively agreeing with them. \r\n\r\nTo address this issue, we introduce **ProsocialDialog**, the first large-scale multi-turn dialogue dataset to teach conversational agents to respond to problematic content following social norms. Covering diverse unethical, problematic, biased, and toxic situations, ProsocialDialog contains responses that encourage prosocial behavior, grounded in commonsense social rules (i.e., rules-of-thumb, RoTs). \r\n\r\n**ProsocialDialog** consists of 58K dialogues between a speaker showing potentially unsafe behavior and a speaker giving constructive feedback for more socially acceptable behavior. Specifically, it contains a rich suite of:\r\n\r\n* 331K utterances\r\n* 160K Rules-of-thumb (RoTs)\r\n* 497K dialogue safety labels accompanied by free-form rationales","description_withheld":null,"homepage":"https://github.com/skywalker023/prosocial-dialog","introduced_date":"2022-05-25","introduced_date_note":null,"introduced_by":{"paper":"/paper/prosocialdialog-a-prosocial-backbone-for","title":"ProsocialDialog: A Prosocial Backbone for Conversational Agents","first_author":"Hyunwoo Kim","url":null},"license":{"name":"CC-BY-4.0","url":"https://github.com/skywalker023/prosocial-dialog/blob/main/LICENSE"},"modalities":[{"name":"Texts","url":"/datasets/modality/texts"},{"name":"Dialog","url":"/datasets/modality/dialog"}],"tasks":[{"name":"Dialogue Understanding","url":"/task/dialogue-understanding","datasets_with_task":"/datasets/task/dialogue-understanding"},{"name":"Dialogue Generation","url":"/task/dialogue-generation","datasets_with_task":"/datasets/task/dialogue-generation"},{"name":"Open-Domain Dialog","url":"/task/open-domain-dialog","datasets_with_task":"/datasets/task/open-domain-dialog"},{"name":"Response Generation","url":"/task/response-generation","datasets_with_task":"/datasets/task/response-generation"},{"name":"Dialogue Safety Prediction","url":"/task/dialogue-safety-prediction","datasets_with_task":"/datasets/task/dialogue-safety-prediction"}],"languages":[{"name":"English","url":"/datasets/language/english"}],"variants":["ProsocialDialog"],"data_loaders":[],"num_papers_in_archive":13,"source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28"},"benchmarks":[{"leaderboard":"/sota/dialogue-safety-prediction-on-prosocialdialog","task":"Dialogue Safety Prediction","dataset_variant":"ProsocialDialog","rows":1,"metrics":["Accuracy"],"first_row_in_archive_order":{"model":"Canary","paper":"/paper/prosocialdialog-a-prosocial-backbone-for","metrics":{"Accuracy":"77.08"},"code_links":[{"title":"skywalker023/prosocial-dialog","url":"https://github.com/skywalker023/prosocial-dialog"}]},"note":"rows are the archive's own order at snapshot; nothing here re-ranks them"}],"papers_with_a_benchmark_row":[{"paper":"/paper/prosocialdialog-a-prosocial-backbone-for","title":"ProsocialDialog: A Prosocial Backbone for Conversational Agents","date":"2022-05-25","rows_on_this_dataset":1,"code_links":1,"syntology":null}],"syntology_totals":{"read_at":"2026-09-24T18:15:14+00:00","papers_with_samples":0,"samples_harvested":0,"samples_ran":0,"samples_unverified":0,"pointer_only_for_licence":0,"papers_with_no_sample_that_ran":0,"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."}