{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:BYVH62TU3EYTYIQ7DOUC2546FO","short_pith_number":"pith:BYVH62TU","schema_version":"1.0","canonical_sha256":"0e2a7f6a74d9313c221f1ba82d779e2ba8c32994ff9e27c2b5256c08db26f885","source":{"kind":"arxiv","id":"2409.12741","version":3},"attestation_state":"computed","paper":{"title":"Fine Tuning Large Language Models for Medicine: The Role and Importance of Direct Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abdessalem Boukil, Ekanath Rangan, Ivan Lopez, Jonathan H Chen, Stephen Ma, Thomas Savage, Vishwesh Patel","submitted_at":"2024-09-19T13:03:24Z","abstract_excerpt":"Large Language Model (LLM) fine tuning is underutilized in the field of medicine. Two of the most common methods of fine tuning are Supervised Fine Tuning (SFT) and Direct Preference Optimization (DPO), but there is little guidance informing users when to use either technique. In this investigation, we compare the performance of SFT and DPO for five common natural language tasks in medicine: Classification with text data, Classification with numeric data, Clinical Reasoning, Summarization, and Clinical Triage. We find that SFT alone is sufficient for Classification with text data, whereas DPO "},"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":"2409.12741","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-09-19T13:03:24Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9bcb2021834e262baa24fc00c517fb8f28c5f66aefa5eb7cac0fb8255eb0b8f7","abstract_canon_sha256":"2c1860dc42bf9e109920c2e01ff1af62f96a9de558e32091ffc7f7a26393ee15"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:48:36.187556Z","signature_b64":"RNGomAj3hi7k4yR3WIbxfsSXO1YF2uzekqPSO0yydaS7HtmTKIKzK7UjqkIMkFMhi7NS2NBHYmNQQk5yIXu9Cw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0e2a7f6a74d9313c221f1ba82d779e2ba8c32994ff9e27c2b5256c08db26f885","last_reissued_at":"2026-07-05T09:48:36.187080Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:48:36.187080Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fine Tuning Large Language Models for Medicine: The Role and Importance of Direct Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Abdessalem Boukil, Ekanath Rangan, Ivan Lopez, Jonathan H Chen, Stephen Ma, Thomas Savage, Vishwesh Patel","submitted_at":"2024-09-19T13:03:24Z","abstract_excerpt":"Large Language Model (LLM) fine tuning is underutilized in the field of medicine. Two of the most common methods of fine tuning are Supervised Fine Tuning (SFT) and Direct Preference Optimization (DPO), but there is little guidance informing users when to use either technique. In this investigation, we compare the performance of SFT and DPO for five common natural language tasks in medicine: Classification with text data, Classification with numeric data, Clinical Reasoning, Summarization, and Clinical Triage. We find that SFT alone is sufficient for Classification with text data, whereas DPO "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.12741","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/2409.12741/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":"2409.12741","created_at":"2026-07-05T09:48:36.187133+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.12741v3","created_at":"2026-07-05T09:48:36.187133+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.12741","created_at":"2026-07-05T09:48:36.187133+00:00"},{"alias_kind":"pith_short_12","alias_value":"BYVH62TU3EYT","created_at":"2026-07-05T09:48:36.187133+00:00"},{"alias_kind":"pith_short_16","alias_value":"BYVH62TU3EYTYIQ7","created_at":"2026-07-05T09:48:36.187133+00:00"},{"alias_kind":"pith_short_8","alias_value":"BYVH62TU","created_at":"2026-07-05T09:48:36.187133+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2605.09584","citing_title":"CLR-voyance: Reinforcing Open-Ended Reasoning for Inpatient Clinical Decision Support with Outcome-Aware Rubrics","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO","json":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO.json","graph_json":"https://pith.science/api/pith-number/BYVH62TU3EYTYIQ7DOUC2546FO/graph.json","events_json":"https://pith.science/api/pith-number/BYVH62TU3EYTYIQ7DOUC2546FO/events.json","paper":"https://pith.science/paper/BYVH62TU"},"agent_actions":{"view_html":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO","download_json":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO.json","view_paper":"https://pith.science/paper/BYVH62TU","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.12741&json=true","fetch_graph":"https://pith.science/api/pith-number/BYVH62TU3EYTYIQ7DOUC2546FO/graph.json","fetch_events":"https://pith.science/api/pith-number/BYVH62TU3EYTYIQ7DOUC2546FO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO/action/storage_attestation","attest_author":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO/action/author_attestation","sign_citation":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO/action/citation_signature","submit_replication":"https://pith.science/pith/BYVH62TU3EYTYIQ7DOUC2546FO/action/replication_record"}},"created_at":"2026-07-05T09:48:36.187133+00:00","updated_at":"2026-07-05T09:48:36.187133+00:00"}