{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:42ZAEA56DVGWKEAM7CKYQFSTPD","short_pith_number":"pith:42ZAEA56","schema_version":"1.0","canonical_sha256":"e6b20203be1d4d65100cf89588165378d6dfbfb6c4b1e1abb6c7e83b03316568","source":{"kind":"arxiv","id":"2503.06011","version":1},"attestation_state":"computed","paper":{"title":"Intent-Aware Self-Correction for Mitigating Social Biases in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Masahiro Kaneko, Naoaki Okazaki, Panatchakorn Anantaprayoon","submitted_at":"2025-03-08T02:20:43Z","abstract_excerpt":"Self-Correction based on feedback improves the output quality of Large Language Models (LLMs). Moreover, as Self-Correction functions like the slow and conscious System-2 thinking from cognitive psychology's perspective, it can potentially reduce LLMs' social biases. LLMs are sensitive to contextual ambiguities and inconsistencies; therefore, explicitly communicating their intentions during interactions when applying Self-Correction for debiasing is crucial. In this study, we demonstrate that clarifying intentions is essential for effectively reducing biases in LLMs through Self-Correction. We"},"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":"2503.06011","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-03-08T02:20:43Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"424e6be573a00614f8686bf736e365f833c3dc2d43c8439aaa14844667125ec9","abstract_canon_sha256":"e8b46dd728d411b615180335dad1ccbf462f2d8ee93e1a0f9b761d895642ec2e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:26:51.577745Z","signature_b64":"wqL7qiVeMO0HZ+/XCBrWnb+zElXHI5gYUYxq5gR995jpHMXEPRpyKaFfwoBvMbQ87NssrXrHkhTkXcOM9/+/Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e6b20203be1d4d65100cf89588165378d6dfbfb6c4b1e1abb6c7e83b03316568","last_reissued_at":"2026-07-05T10:26:51.577261Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:26:51.577261Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Intent-Aware Self-Correction for Mitigating Social Biases in Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Masahiro Kaneko, Naoaki Okazaki, Panatchakorn Anantaprayoon","submitted_at":"2025-03-08T02:20:43Z","abstract_excerpt":"Self-Correction based on feedback improves the output quality of Large Language Models (LLMs). Moreover, as Self-Correction functions like the slow and conscious System-2 thinking from cognitive psychology's perspective, it can potentially reduce LLMs' social biases. LLMs are sensitive to contextual ambiguities and inconsistencies; therefore, explicitly communicating their intentions during interactions when applying Self-Correction for debiasing is crucial. In this study, we demonstrate that clarifying intentions is essential for effectively reducing biases in LLMs through Self-Correction. We"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.06011","kind":"arxiv","version":1},"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/2503.06011/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":"2503.06011","created_at":"2026-07-05T10:26:51.577324+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.06011v1","created_at":"2026-07-05T10:26:51.577324+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.06011","created_at":"2026-07-05T10:26:51.577324+00:00"},{"alias_kind":"pith_short_12","alias_value":"42ZAEA56DVGW","created_at":"2026-07-05T10:26:51.577324+00:00"},{"alias_kind":"pith_short_16","alias_value":"42ZAEA56DVGWKEAM","created_at":"2026-07-05T10:26:51.577324+00:00"},{"alias_kind":"pith_short_8","alias_value":"42ZAEA56","created_at":"2026-07-05T10:26:51.577324+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.30989","citing_title":"Wait, am I Being Fair? Characterizing Deductive Stereotyping and Mitigating It with Fair-GCG","ref_index":85,"is_internal_anchor":false},{"citing_arxiv_id":"2510.17062","citing_title":"Investigating Thinking Behaviours of Reasoning-Based Language Models for Social Bias Mitigation","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD","json":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD.json","graph_json":"https://pith.science/api/pith-number/42ZAEA56DVGWKEAM7CKYQFSTPD/graph.json","events_json":"https://pith.science/api/pith-number/42ZAEA56DVGWKEAM7CKYQFSTPD/events.json","paper":"https://pith.science/paper/42ZAEA56"},"agent_actions":{"view_html":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD","download_json":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD.json","view_paper":"https://pith.science/paper/42ZAEA56","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.06011&json=true","fetch_graph":"https://pith.science/api/pith-number/42ZAEA56DVGWKEAM7CKYQFSTPD/graph.json","fetch_events":"https://pith.science/api/pith-number/42ZAEA56DVGWKEAM7CKYQFSTPD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD/action/storage_attestation","attest_author":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD/action/author_attestation","sign_citation":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD/action/citation_signature","submit_replication":"https://pith.science/pith/42ZAEA56DVGWKEAM7CKYQFSTPD/action/replication_record"}},"created_at":"2026-07-05T10:26:51.577324+00:00","updated_at":"2026-07-05T10:26:51.577324+00:00"}