{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:7UVS65UNMXQVABZDIGXUWUWDHX","short_pith_number":"pith:7UVS65UN","schema_version":"1.0","canonical_sha256":"fd2b2f768d65e150072341af4b52c33dff94c88546d422bc0b29be5159f6a10e","source":{"kind":"arxiv","id":"2606.05463","version":1},"attestation_state":"computed","paper":{"title":"PSEBench: A Controllable and Verifiable Benchmark for Evaluating LLMs in Patient Safety Event Triage","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Annabel Strauss, Carl Yang, Che Ngufor, Katharine M. Nesbitt, Keqi Han, Lindsey Hughes, Nicole Schueler, Ryan Young, Yuan Xue, Zhijun Yin","submitted_at":"2026-06-03T21:41:39Z","abstract_excerpt":"Patient safety event triage, determining whether a clinical event is reportable under jurisdiction-specific policy, is a high-stakes task typically performed manually by patient safety experts. Although LLMs may support this workflow, reliable evaluation is limited by the lack of benchmarks to capture evidence-grounded policy reasoning, proactive information seeking for incomplete reports, and principled abstention in irreducibly ambiguous cases. We address this gap with a policy-grounded construction methodology centered on the clause card, a structured representation that factorizes regulato"},"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":"2606.05463","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-03T21:41:39Z","cross_cats_sorted":[],"title_canon_sha256":"c36f18df71a9e517ea240cb488346c8df5f3e8cc08c7be079243fede2a15f468","abstract_canon_sha256":"bdaa00ea7d7f33684083dcfd09d9f89cbea032475d36e1aadad1c3ce77a2d3f5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-05T01:14:51.909138Z","signature_b64":"csxMPuAFHhPBwu+RNHYvudXYPifFooOgc2pntFll/H6T+lrn32WGSEksva0KTD5P7rCfRT7eLM7mirLUbidDDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fd2b2f768d65e150072341af4b52c33dff94c88546d422bc0b29be5159f6a10e","last_reissued_at":"2026-06-05T01:14:51.908651Z","signature_status":"signed_v1","first_computed_at":"2026-06-05T01:14:51.908651Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PSEBench: A Controllable and Verifiable Benchmark for Evaluating LLMs in Patient Safety Event Triage","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Annabel Strauss, Carl Yang, Che Ngufor, Katharine M. Nesbitt, Keqi Han, Lindsey Hughes, Nicole Schueler, Ryan Young, Yuan Xue, Zhijun Yin","submitted_at":"2026-06-03T21:41:39Z","abstract_excerpt":"Patient safety event triage, determining whether a clinical event is reportable under jurisdiction-specific policy, is a high-stakes task typically performed manually by patient safety experts. Although LLMs may support this workflow, reliable evaluation is limited by the lack of benchmarks to capture evidence-grounded policy reasoning, proactive information seeking for incomplete reports, and principled abstention in irreducibly ambiguous cases. We address this gap with a policy-grounded construction methodology centered on the clause card, a structured representation that factorizes regulato"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.05463","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/2606.05463/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":"2606.05463","created_at":"2026-06-05T01:14:51.908710+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.05463v1","created_at":"2026-06-05T01:14:51.908710+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.05463","created_at":"2026-06-05T01:14:51.908710+00:00"},{"alias_kind":"pith_short_12","alias_value":"7UVS65UNMXQV","created_at":"2026-06-05T01:14:51.908710+00:00"},{"alias_kind":"pith_short_16","alias_value":"7UVS65UNMXQVABZD","created_at":"2026-06-05T01:14:51.908710+00:00"},{"alias_kind":"pith_short_8","alias_value":"7UVS65UN","created_at":"2026-06-05T01:14:51.908710+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/7UVS65UNMXQVABZDIGXUWUWDHX","json":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX.json","graph_json":"https://pith.science/api/pith-number/7UVS65UNMXQVABZDIGXUWUWDHX/graph.json","events_json":"https://pith.science/api/pith-number/7UVS65UNMXQVABZDIGXUWUWDHX/events.json","paper":"https://pith.science/paper/7UVS65UN"},"agent_actions":{"view_html":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX","download_json":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX.json","view_paper":"https://pith.science/paper/7UVS65UN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.05463&json=true","fetch_graph":"https://pith.science/api/pith-number/7UVS65UNMXQVABZDIGXUWUWDHX/graph.json","fetch_events":"https://pith.science/api/pith-number/7UVS65UNMXQVABZDIGXUWUWDHX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX/action/storage_attestation","attest_author":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX/action/author_attestation","sign_citation":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX/action/citation_signature","submit_replication":"https://pith.science/pith/7UVS65UNMXQVABZDIGXUWUWDHX/action/replication_record"}},"created_at":"2026-06-05T01:14:51.908710+00:00","updated_at":"2026-06-05T01:14:51.908710+00:00"}