{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2021:IPDSZVJS3DWOWQMKHSCQUJH66A","short_pith_number":"pith:IPDSZVJS","schema_version":"1.0","canonical_sha256":"43c72cd532d8eceb418a3c850a24fef0013173802d8c8b919edc6ea6b1edac43","source":{"kind":"arxiv","id":"2109.08914","version":2},"attestation_state":"computed","paper":{"title":"Text Detoxification using Large Pre-trained Neural Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alexander Panchenko, Anton Voronov, Daryna Dementieva, David Dale, Nikita Semenov, Olga Kozlova, Varvara Logacheva","submitted_at":"2021-09-18T11:55:32Z","abstract_excerpt":"We present two novel unsupervised methods for eliminating toxicity in text. Our first method combines two recent ideas: (1) guidance of the generation process with small style-conditional language models and (2) use of paraphrasing models to perform style transfer. We use a well-performing paraphraser guided by style-trained language models to keep the text content and remove toxicity. Our second method uses BERT to replace toxic words with their non-offensive synonyms. We make the method more flexible by enabling BERT to replace mask tokens with a variable number of words. Finally, we present"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2109.08914","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2021-09-18T11:55:32Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"bb2560aa9f02677b73cd042df7a99e0e9d325f8ab7fba3aea495d47224885eba","abstract_canon_sha256":"eb4baf71b05e312054e9d07ffc8a4bab980a5dde7bd59e4ac27551efa8244934"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T03:28:41.063132Z","signature_b64":"lpX5ncaCf5kefmwQj6WHVoAr/vc4bQC7BU75pRUw5jQhO4n3YAD9gNLiDRqXjOgkF+BBST+DgfRtHEtaqjBEDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"43c72cd532d8eceb418a3c850a24fef0013173802d8c8b919edc6ea6b1edac43","last_reissued_at":"2026-07-05T03:28:41.062683Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T03:28:41.062683Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Text Detoxification using Large Pre-trained Neural Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Alexander Panchenko, Anton Voronov, Daryna Dementieva, David Dale, Nikita Semenov, Olga Kozlova, Varvara Logacheva","submitted_at":"2021-09-18T11:55:32Z","abstract_excerpt":"We present two novel unsupervised methods for eliminating toxicity in text. Our first method combines two recent ideas: (1) guidance of the generation process with small style-conditional language models and (2) use of paraphrasing models to perform style transfer. We use a well-performing paraphraser guided by style-trained language models to keep the text content and remove toxicity. Our second method uses BERT to replace toxic words with their non-offensive synonyms. We make the method more flexible by enabling BERT to replace mask tokens with a variable number of words. Finally, we present"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2109.08914","kind":"arxiv","version":2},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2109.08914/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2109.08914","created_at":"2026-07-05T03:28:41.062738+00:00"},{"alias_kind":"arxiv_version","alias_value":"2109.08914v2","created_at":"2026-07-05T03:28:41.062738+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2109.08914","created_at":"2026-07-05T03:28:41.062738+00:00"},{"alias_kind":"pith_short_12","alias_value":"IPDSZVJS3DWO","created_at":"2026-07-05T03:28:41.062738+00:00"},{"alias_kind":"pith_short_16","alias_value":"IPDSZVJS3DWOWQMK","created_at":"2026-07-05T03:28:41.062738+00:00"},{"alias_kind":"pith_short_8","alias_value":"IPDSZVJS","created_at":"2026-07-05T03:28:41.062738+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A","json":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A.json","graph_json":"https://pith.science/api/pith-number/IPDSZVJS3DWOWQMKHSCQUJH66A/graph.json","events_json":"https://pith.science/api/pith-number/IPDSZVJS3DWOWQMKHSCQUJH66A/events.json","paper":"https://pith.science/paper/IPDSZVJS"},"agent_actions":{"view_html":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A","download_json":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A.json","view_paper":"https://pith.science/paper/IPDSZVJS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2109.08914&json=true","fetch_graph":"https://pith.science/api/pith-number/IPDSZVJS3DWOWQMKHSCQUJH66A/graph.json","fetch_events":"https://pith.science/api/pith-number/IPDSZVJS3DWOWQMKHSCQUJH66A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A/action/storage_attestation","attest_author":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A/action/author_attestation","sign_citation":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A/action/citation_signature","submit_replication":"https://pith.science/pith/IPDSZVJS3DWOWQMKHSCQUJH66A/action/replication_record"}},"created_at":"2026-07-05T03:28:41.062738+00:00","updated_at":"2026-07-05T03:28:41.062738+00:00"}