{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:IQGMMGWCCZQIPBVD37BFHNUDDR","short_pith_number":"pith:IQGMMGWC","schema_version":"1.0","canonical_sha256":"440cc61ac216608786a3dfc253b6831c728855c9d755550b9868c0a2a2b9d7c6","source":{"kind":"arxiv","id":"2307.06439","version":1},"attestation_state":"computed","paper":{"title":"Distilling Large Language Models for Biomedical Knowledge Extraction: A Case Study on Adverse Drug Events","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Cliff Wong, Erika Strandberg, Hoifung Poon, Mu Wei, Naoto Usuyama, Naveen Valluri, Praneeth Sanapathi, Sheng Zhang, Tristan Naumann, Yonas Woldesenbet, Yu Gu","submitted_at":"2023-07-12T20:08:48Z","abstract_excerpt":"Large language models (LLMs), such as GPT-4, have demonstrated remarkable capabilities across a wide range of tasks, including health applications. In this paper, we study how LLMs can be used to scale biomedical knowledge curation. We find that while LLMs already possess decent competency in structuring biomedical text, by distillation into a task-specific student model through self-supervised learning, substantial gains can be attained over out-of-box LLMs, with additional advantages such as cost, efficiency, and white-box model access.\n  We conduct a case study on adverse drug event (ADE) e"},"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":"2307.06439","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-07-12T20:08:48Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8fa4227da43a6bd474383c4e9c46ce1641be97aa74c993bedc96bba6248da1cf","abstract_canon_sha256":"f712806efe8df4f8a877d99aaab90cf751ce926800d07f077a68990543f53c6d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:30:33.846347Z","signature_b64":"oc2zjI+PP5g9j4Lf613uL9k2XsbXghmXK/seuAPn10XfnVys81uGN+N3B7ekFdY8fXjruNJsHXNJJX7NQwhiDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"440cc61ac216608786a3dfc253b6831c728855c9d755550b9868c0a2a2b9d7c6","last_reissued_at":"2026-07-05T06:30:33.845914Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:30:33.845914Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Distilling Large Language Models for Biomedical Knowledge Extraction: A Case Study on Adverse Drug Events","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Cliff Wong, Erika Strandberg, Hoifung Poon, Mu Wei, Naoto Usuyama, Naveen Valluri, Praneeth Sanapathi, Sheng Zhang, Tristan Naumann, Yonas Woldesenbet, Yu Gu","submitted_at":"2023-07-12T20:08:48Z","abstract_excerpt":"Large language models (LLMs), such as GPT-4, have demonstrated remarkable capabilities across a wide range of tasks, including health applications. In this paper, we study how LLMs can be used to scale biomedical knowledge curation. We find that while LLMs already possess decent competency in structuring biomedical text, by distillation into a task-specific student model through self-supervised learning, substantial gains can be attained over out-of-box LLMs, with additional advantages such as cost, efficiency, and white-box model access.\n  We conduct a case study on adverse drug event (ADE) e"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.06439","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/2307.06439/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":"2307.06439","created_at":"2026-07-05T06:30:33.845986+00:00"},{"alias_kind":"arxiv_version","alias_value":"2307.06439v1","created_at":"2026-07-05T06:30:33.845986+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.06439","created_at":"2026-07-05T06:30:33.845986+00:00"},{"alias_kind":"pith_short_12","alias_value":"IQGMMGWCCZQI","created_at":"2026-07-05T06:30:33.845986+00:00"},{"alias_kind":"pith_short_16","alias_value":"IQGMMGWCCZQIPBVD","created_at":"2026-07-05T06:30:33.845986+00:00"},{"alias_kind":"pith_short_8","alias_value":"IQGMMGWC","created_at":"2026-07-05T06:30:33.845986+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.08268","citing_title":"Different Teachers, Different Capabilities: Sub-1B On-Device Distillation for Structured Text Enrichment","ref_index":9,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR","json":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR.json","graph_json":"https://pith.science/api/pith-number/IQGMMGWCCZQIPBVD37BFHNUDDR/graph.json","events_json":"https://pith.science/api/pith-number/IQGMMGWCCZQIPBVD37BFHNUDDR/events.json","paper":"https://pith.science/paper/IQGMMGWC"},"agent_actions":{"view_html":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR","download_json":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR.json","view_paper":"https://pith.science/paper/IQGMMGWC","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2307.06439&json=true","fetch_graph":"https://pith.science/api/pith-number/IQGMMGWCCZQIPBVD37BFHNUDDR/graph.json","fetch_events":"https://pith.science/api/pith-number/IQGMMGWCCZQIPBVD37BFHNUDDR/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR/action/storage_attestation","attest_author":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR/action/author_attestation","sign_citation":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR/action/citation_signature","submit_replication":"https://pith.science/pith/IQGMMGWCCZQIPBVD37BFHNUDDR/action/replication_record"}},"created_at":"2026-07-05T06:30:33.845986+00:00","updated_at":"2026-07-05T06:30:33.845986+00:00"}