{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:KDBLOAXAXKOA3R4NIZQXLYUEA3","short_pith_number":"pith:KDBLOAXA","schema_version":"1.0","canonical_sha256":"50c2b702e0ba9c0dc78d466175e28406d514064d5af7645d093194bd32b6828c","source":{"kind":"arxiv","id":"2302.00871","version":3},"attestation_state":"computed","paper":{"title":"Using In-Context Learning to Improve Dialogue Safety","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Devamanyu Hazarika, Di Jin, Dilek Hakkani-T\\\"ur, Nicholas Meade, Prakhar Gupta, Siva Reddy, Spandana Gella, Yang Liu","submitted_at":"2023-02-02T04:46:03Z","abstract_excerpt":"While large neural-based conversational models have become increasingly proficient dialogue agents, recent work has highlighted safety issues with these systems. For example, these systems can be goaded into generating toxic content, which often perpetuates social biases or stereotypes. We investigate a retrieval-based method for reducing bias and toxicity in responses from chatbots. It uses in-context learning to steer a model towards safer generations. Concretely, to generate a response to an unsafe dialogue context, we retrieve demonstrations of safe responses to similar dialogue contexts. "},"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":"2302.00871","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-02-02T04:46:03Z","cross_cats_sorted":[],"title_canon_sha256":"0b40c0c2bf6561a1d8047cdb1bf7009aea838ad1c64eb232ad84b2f3d346990f","abstract_canon_sha256":"db0df38b6215cac1d53294602eb2279194408805e02743edfb028c9f1a867b21"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:03:20.432356Z","signature_b64":"2zbKDPdx5X1bacofClLQgdBPu2wDQvJVrsSk4D3hNV72U1xVsPEVvcQZq1sZ5rw5rQvbWORcnyeg7ic6LUukCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"50c2b702e0ba9c0dc78d466175e28406d514064d5af7645d093194bd32b6828c","last_reissued_at":"2026-07-05T07:03:20.431867Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:03:20.431867Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Using In-Context Learning to Improve Dialogue Safety","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Devamanyu Hazarika, Di Jin, Dilek Hakkani-T\\\"ur, Nicholas Meade, Prakhar Gupta, Siva Reddy, Spandana Gella, Yang Liu","submitted_at":"2023-02-02T04:46:03Z","abstract_excerpt":"While large neural-based conversational models have become increasingly proficient dialogue agents, recent work has highlighted safety issues with these systems. For example, these systems can be goaded into generating toxic content, which often perpetuates social biases or stereotypes. We investigate a retrieval-based method for reducing bias and toxicity in responses from chatbots. It uses in-context learning to steer a model towards safer generations. Concretely, to generate a response to an unsafe dialogue context, we retrieve demonstrations of safe responses to similar dialogue contexts. "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2302.00871","kind":"arxiv","version":3},"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/2302.00871/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":"2302.00871","created_at":"2026-07-05T07:03:20.431923+00:00"},{"alias_kind":"arxiv_version","alias_value":"2302.00871v3","created_at":"2026-07-05T07:03:20.431923+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2302.00871","created_at":"2026-07-05T07:03:20.431923+00:00"},{"alias_kind":"pith_short_12","alias_value":"KDBLOAXAXKOA","created_at":"2026-07-05T07:03:20.431923+00:00"},{"alias_kind":"pith_short_16","alias_value":"KDBLOAXAXKOA3R4N","created_at":"2026-07-05T07:03:20.431923+00:00"},{"alias_kind":"pith_short_8","alias_value":"KDBLOAXA","created_at":"2026-07-05T07:03:20.431923+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.05552","citing_title":"Designing with Tensions: Older Adults' Emotional Support-Seeking Under System-Level Constraints in Conversational AI","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2305.06161","citing_title":"StarCoder: may the source be with you!","ref_index":70,"is_internal_anchor":false},{"citing_arxiv_id":"2305.06161","citing_title":"StarCoder: may the source be with you!","ref_index":274,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3","json":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3.json","graph_json":"https://pith.science/api/pith-number/KDBLOAXAXKOA3R4NIZQXLYUEA3/graph.json","events_json":"https://pith.science/api/pith-number/KDBLOAXAXKOA3R4NIZQXLYUEA3/events.json","paper":"https://pith.science/paper/KDBLOAXA"},"agent_actions":{"view_html":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3","download_json":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3.json","view_paper":"https://pith.science/paper/KDBLOAXA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2302.00871&json=true","fetch_graph":"https://pith.science/api/pith-number/KDBLOAXAXKOA3R4NIZQXLYUEA3/graph.json","fetch_events":"https://pith.science/api/pith-number/KDBLOAXAXKOA3R4NIZQXLYUEA3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3/action/storage_attestation","attest_author":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3/action/author_attestation","sign_citation":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3/action/citation_signature","submit_replication":"https://pith.science/pith/KDBLOAXAXKOA3R4NIZQXLYUEA3/action/replication_record"}},"created_at":"2026-07-05T07:03:20.431923+00:00","updated_at":"2026-07-05T07:03:20.431923+00:00"}