{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:2HAUTEMJUMRPLPDKTSZM6PAGUG","short_pith_number":"pith:2HAUTEMJ","schema_version":"1.0","canonical_sha256":"d1c1499189a322f5bc6a9cb2cf3c06a19f922fa1886328c989b7f81f27db4906","source":{"kind":"arxiv","id":"2606.27709","version":1},"attestation_state":"computed","paper":{"title":"Low-Agreeableness Persona Conditioning for Safe LLM Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Austin MY Cheung, Yi Yang","submitted_at":"2026-06-26T04:15:06Z","abstract_excerpt":"Recent work has shown that fine-tuning large language models (LLMs) for social warmth degrades factual reliability and increases sycophancy. We investigate a related but distinct failure mode: warmth fine-tuning also weakens adversarial safety, making models more susceptible to jailbreaks and harmful output generation. We examine whether this reflects an inherent consequence of empathetic adaptation or an artifact of data construction. To address this, we introduce a persona-driven rewriting pipeline that conditions user turns on low agreeableness and pairs this with warm, de-escalating assist"},"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":"2606.27709","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-06-26T04:15:06Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"484d12647ce027a17e34659094bf3b21070ac29215fc2b83220ceb36aadf5696","abstract_canon_sha256":"a8ae036aafbcdf96c3ae6f61d9c959cc6b45acb0a135c8901b515145eb13c87a"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T01:14:46.061401Z","signature_b64":"zwNQ/l0ywyvF8KDLVcnzy1QnTRdN/uyXsKc535dI7X+FBV8uW2gXIWK7AnRzOMGD6mW1kxH/TLShPV/avPjYCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d1c1499189a322f5bc6a9cb2cf3c06a19f922fa1886328c989b7f81f27db4906","last_reissued_at":"2026-06-29T01:14:46.061046Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T01:14:46.061046Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Low-Agreeableness Persona Conditioning for Safe LLM Fine-Tuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Austin MY Cheung, Yi Yang","submitted_at":"2026-06-26T04:15:06Z","abstract_excerpt":"Recent work has shown that fine-tuning large language models (LLMs) for social warmth degrades factual reliability and increases sycophancy. We investigate a related but distinct failure mode: warmth fine-tuning also weakens adversarial safety, making models more susceptible to jailbreaks and harmful output generation. We examine whether this reflects an inherent consequence of empathetic adaptation or an artifact of data construction. To address this, we introduce a persona-driven rewriting pipeline that conditions user turns on low agreeableness and pairs this with warm, de-escalating assist"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27709","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/2606.27709/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":"2606.27709","created_at":"2026-06-29T01:14:46.061094+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.27709v1","created_at":"2026-06-29T01:14:46.061094+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27709","created_at":"2026-06-29T01:14:46.061094+00:00"},{"alias_kind":"pith_short_12","alias_value":"2HAUTEMJUMRP","created_at":"2026-06-29T01:14:46.061094+00:00"},{"alias_kind":"pith_short_16","alias_value":"2HAUTEMJUMRPLPDK","created_at":"2026-06-29T01:14:46.061094+00:00"},{"alias_kind":"pith_short_8","alias_value":"2HAUTEMJ","created_at":"2026-06-29T01:14:46.061094+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/2HAUTEMJUMRPLPDKTSZM6PAGUG","json":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG.json","graph_json":"https://pith.science/api/pith-number/2HAUTEMJUMRPLPDKTSZM6PAGUG/graph.json","events_json":"https://pith.science/api/pith-number/2HAUTEMJUMRPLPDKTSZM6PAGUG/events.json","paper":"https://pith.science/paper/2HAUTEMJ"},"agent_actions":{"view_html":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG","download_json":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG.json","view_paper":"https://pith.science/paper/2HAUTEMJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.27709&json=true","fetch_graph":"https://pith.science/api/pith-number/2HAUTEMJUMRPLPDKTSZM6PAGUG/graph.json","fetch_events":"https://pith.science/api/pith-number/2HAUTEMJUMRPLPDKTSZM6PAGUG/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG/action/timestamp_anchor","attest_storage":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG/action/storage_attestation","attest_author":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG/action/author_attestation","sign_citation":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG/action/citation_signature","submit_replication":"https://pith.science/pith/2HAUTEMJUMRPLPDKTSZM6PAGUG/action/replication_record"}},"created_at":"2026-06-29T01:14:46.061094+00:00","updated_at":"2026-06-29T01:14:46.061094+00:00"}