{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:DXG3ARNIC6KNPLJUYRNL5M46D6","short_pith_number":"pith:DXG3ARNI","schema_version":"1.0","canonical_sha256":"1dcdb045a81794d7ad34c45abeb39e1fa424725e1f3042a0ee847cb0797b94db","source":{"kind":"arxiv","id":"2402.11764","version":2},"attestation_state":"computed","paper":{"title":"ChatGPT Based Data Augmentation for Improved Parameter-Efficient Debiasing of LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CY"],"primary_cat":"cs.CL","authors_text":"Adhithya Saravanan, Anima Anandkumar, Or Sharir, Pengrui Han, Rafal Kocielnik, Roy Jiang","submitted_at":"2024-02-19T01:28:48Z","abstract_excerpt":"Large Language models (LLMs), while powerful, exhibit harmful social biases. Debiasing is often challenging due to computational costs, data constraints, and potential degradation of multi-task language capabilities. This work introduces a novel approach utilizing ChatGPT to generate synthetic training data, aiming to enhance the debiasing of LLMs. We propose two strategies: Targeted Prompting, which provides effective debiasing for known biases but necessitates prior specification of bias in question; and General Prompting, which, while slightly less effective, offers debiasing across various"},"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":"2402.11764","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-02-19T01:28:48Z","cross_cats_sorted":["cs.AI","cs.CY"],"title_canon_sha256":"36f89dab5af94cf1e93baebf41293e34449a0170e29eb26253d62e351c5b7968","abstract_canon_sha256":"8c33abc2a8116143954fadf5585ae123f28b0832ee627dab6a0238f706efb989"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:07:26.437104Z","signature_b64":"ri4sHIRfZyihKyGVCapsZEweaSet0dUD4/4rKXz7UkoAjyIoCkouzc/NV0N67wxhOY79m1JGZdg5CvUrxJz/Bw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"1dcdb045a81794d7ad34c45abeb39e1fa424725e1f3042a0ee847cb0797b94db","last_reissued_at":"2026-07-05T09:07:26.436673Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:07:26.436673Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ChatGPT Based Data Augmentation for Improved Parameter-Efficient Debiasing of LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CY"],"primary_cat":"cs.CL","authors_text":"Adhithya Saravanan, Anima Anandkumar, Or Sharir, Pengrui Han, Rafal Kocielnik, Roy Jiang","submitted_at":"2024-02-19T01:28:48Z","abstract_excerpt":"Large Language models (LLMs), while powerful, exhibit harmful social biases. Debiasing is often challenging due to computational costs, data constraints, and potential degradation of multi-task language capabilities. This work introduces a novel approach utilizing ChatGPT to generate synthetic training data, aiming to enhance the debiasing of LLMs. We propose two strategies: Targeted Prompting, which provides effective debiasing for known biases but necessitates prior specification of bias in question; and General Prompting, which, while slightly less effective, offers debiasing across various"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.11764","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/2402.11764/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":"2402.11764","created_at":"2026-07-05T09:07:26.436738+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.11764v2","created_at":"2026-07-05T09:07:26.436738+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.11764","created_at":"2026-07-05T09:07:26.436738+00:00"},{"alias_kind":"pith_short_12","alias_value":"DXG3ARNIC6KN","created_at":"2026-07-05T09:07:26.436738+00:00"},{"alias_kind":"pith_short_16","alias_value":"DXG3ARNIC6KNPLJU","created_at":"2026-07-05T09:07:26.436738+00:00"},{"alias_kind":"pith_short_8","alias_value":"DXG3ARNI","created_at":"2026-07-05T09:07:26.436738+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.12730","citing_title":"Rethinking Psychometric Evaluation of LLMs: When and Why Self-Reports Predict Behavior","ref_index":46,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6","json":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6.json","graph_json":"https://pith.science/api/pith-number/DXG3ARNIC6KNPLJUYRNL5M46D6/graph.json","events_json":"https://pith.science/api/pith-number/DXG3ARNIC6KNPLJUYRNL5M46D6/events.json","paper":"https://pith.science/paper/DXG3ARNI"},"agent_actions":{"view_html":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6","download_json":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6.json","view_paper":"https://pith.science/paper/DXG3ARNI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.11764&json=true","fetch_graph":"https://pith.science/api/pith-number/DXG3ARNIC6KNPLJUYRNL5M46D6/graph.json","fetch_events":"https://pith.science/api/pith-number/DXG3ARNIC6KNPLJUYRNL5M46D6/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6/action/timestamp_anchor","attest_storage":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6/action/storage_attestation","attest_author":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6/action/author_attestation","sign_citation":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6/action/citation_signature","submit_replication":"https://pith.science/pith/DXG3ARNIC6KNPLJUYRNL5M46D6/action/replication_record"}},"created_at":"2026-07-05T09:07:26.436738+00:00","updated_at":"2026-07-05T09:07:26.436738+00:00"}