{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:FOXU3NHWA36VPSRGATTK5FMI24","short_pith_number":"pith:FOXU3NHW","schema_version":"1.0","canonical_sha256":"2baf4db4f606fd57ca2604e6ae9588d738cd01ce963c2cb4253a6b257fcef8d1","source":{"kind":"arxiv","id":"2412.04573","version":1},"attestation_state":"computed","paper":{"title":"Give me Some Hard Questions: Synthetic Data Generation for Clinical QA","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ardavan Saeedi, Fan Bai, Hamid Hassanzadeh, Joel Stremmel, Keith Harrigian, Mark Dredze","submitted_at":"2024-12-05T19:35:41Z","abstract_excerpt":"Clinical Question Answering (QA) systems enable doctors to quickly access patient information from electronic health records (EHRs). However, training these systems requires significant annotated data, which is limited due to the expertise needed and the privacy concerns associated with clinical data. This paper explores generating Clinical QA data using large language models (LLMs) in a zero-shot setting. We find that naive prompting often results in easy questions that do not reflect the complexity of clinical scenarios. To address this, we propose two prompting strategies: 1) instructing th"},"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":"2412.04573","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-05T19:35:41Z","cross_cats_sorted":[],"title_canon_sha256":"d2c3dfc94de5fea17fa593321cc77fc29444a543fed219883e3c71a9fba862ed","abstract_canon_sha256":"ec8d91993906e76bcf6199b7152f843d61632613687b651a8ffcac60dee8f5c0"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:45:20.225971Z","signature_b64":"QHUIwEuXQzDuD/oeLWFaKOqa1+JupMKpn+LHFDCCV5VMhp9uE2xFWeLUKjiclShc0Dvh8qJY7b74TO30kif4BQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2baf4db4f606fd57ca2604e6ae9588d738cd01ce963c2cb4253a6b257fcef8d1","last_reissued_at":"2026-07-05T09:45:20.225448Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:45:20.225448Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Give me Some Hard Questions: Synthetic Data Generation for Clinical QA","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Ardavan Saeedi, Fan Bai, Hamid Hassanzadeh, Joel Stremmel, Keith Harrigian, Mark Dredze","submitted_at":"2024-12-05T19:35:41Z","abstract_excerpt":"Clinical Question Answering (QA) systems enable doctors to quickly access patient information from electronic health records (EHRs). However, training these systems requires significant annotated data, which is limited due to the expertise needed and the privacy concerns associated with clinical data. This paper explores generating Clinical QA data using large language models (LLMs) in a zero-shot setting. We find that naive prompting often results in easy questions that do not reflect the complexity of clinical scenarios. To address this, we propose two prompting strategies: 1) instructing th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.04573","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/2412.04573/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":"2412.04573","created_at":"2026-07-05T09:45:20.225505+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.04573v1","created_at":"2026-07-05T09:45:20.225505+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.04573","created_at":"2026-07-05T09:45:20.225505+00:00"},{"alias_kind":"pith_short_12","alias_value":"FOXU3NHWA36V","created_at":"2026-07-05T09:45:20.225505+00:00"},{"alias_kind":"pith_short_16","alias_value":"FOXU3NHWA36VPSRG","created_at":"2026-07-05T09:45:20.225505+00:00"},{"alias_kind":"pith_short_8","alias_value":"FOXU3NHW","created_at":"2026-07-05T09:45:20.225505+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2505.15024","citing_title":"Diagnosing our datasets: How does my language model learn clinical information?","ref_index":5,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24","json":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24.json","graph_json":"https://pith.science/api/pith-number/FOXU3NHWA36VPSRGATTK5FMI24/graph.json","events_json":"https://pith.science/api/pith-number/FOXU3NHWA36VPSRGATTK5FMI24/events.json","paper":"https://pith.science/paper/FOXU3NHW"},"agent_actions":{"view_html":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24","download_json":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24.json","view_paper":"https://pith.science/paper/FOXU3NHW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.04573&json=true","fetch_graph":"https://pith.science/api/pith-number/FOXU3NHWA36VPSRGATTK5FMI24/graph.json","fetch_events":"https://pith.science/api/pith-number/FOXU3NHWA36VPSRGATTK5FMI24/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24/action/storage_attestation","attest_author":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24/action/author_attestation","sign_citation":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24/action/citation_signature","submit_replication":"https://pith.science/pith/FOXU3NHWA36VPSRGATTK5FMI24/action/replication_record"}},"created_at":"2026-07-05T09:45:20.225505+00:00","updated_at":"2026-07-05T09:45:20.225505+00:00"}