{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:MTZKXHAXDGJPFJHYOFLRESQMK3","short_pith_number":"pith:MTZKXHAX","schema_version":"1.0","canonical_sha256":"64f2ab9c171992f2a4f87157124a0c56c409af7583cc0b45d43c7c101408d127","source":{"kind":"arxiv","id":"2601.17642","version":2},"attestation_state":"computed","paper":{"title":"Health-ORSC-Bench: A Benchmark for Measuring Over-Refusal and Safety Completion in Health Context","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Guanghao Wu, Heng Ji, Liting Huang, Preslav Nakov, Usman Naseem, Zhihao Zhang","submitted_at":"2026-01-25T01:28:52Z","abstract_excerpt":"Safety alignment in Large Language Models is critical for healthcare; however, reliance on binary refusal boundaries often results in over-refusal of benign queries or unsafe compliance with harmful ones. While existing benchmarks measure these extremes, they fail to evaluate Safe Completion: the model's ability to maximise helpfulness on dual-use or borderline queries by providing safe, high-level guidance without crossing into actionable harm. We introduce Health-ORSC-Bench, the first large-scale benchmark designed to systematically measure Over-Refusal and Safe Completion quality in healthc"},"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":"2601.17642","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.AI","submitted_at":"2026-01-25T01:28:52Z","cross_cats_sorted":[],"title_canon_sha256":"7fa5290c2dc425fa66ded678658b23bdef2a58e0c7d1657d991b5f8bd5f77da1","abstract_canon_sha256":"c5496972866afb55c3e9f9104b093617abf6dc86d2c131d97897b19e4a1b3bcb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-29T01:14:28.092579Z","signature_b64":"r8/WYHu4sfVb+n2xRpYkUJiMYWfJhkH2aDAyKOU1OS2yi2JU7l//OeOzwsWN+yq/1yL0ji6E7OqKga0Wv1SBAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64f2ab9c171992f2a4f87157124a0c56c409af7583cc0b45d43c7c101408d127","last_reissued_at":"2026-06-29T01:14:28.091971Z","signature_status":"signed_v1","first_computed_at":"2026-06-29T01:14:28.091971Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Health-ORSC-Bench: A Benchmark for Measuring Over-Refusal and Safety Completion in Health Context","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Guanghao Wu, Heng Ji, Liting Huang, Preslav Nakov, Usman Naseem, Zhihao Zhang","submitted_at":"2026-01-25T01:28:52Z","abstract_excerpt":"Safety alignment in Large Language Models is critical for healthcare; however, reliance on binary refusal boundaries often results in over-refusal of benign queries or unsafe compliance with harmful ones. While existing benchmarks measure these extremes, they fail to evaluate Safe Completion: the model's ability to maximise helpfulness on dual-use or borderline queries by providing safe, high-level guidance without crossing into actionable harm. We introduce Health-ORSC-Bench, the first large-scale benchmark designed to systematically measure Over-Refusal and Safe Completion quality in healthc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.17642","kind":"arxiv","version":2},"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/2601.17642/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":"2601.17642","created_at":"2026-06-29T01:14:28.092041+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.17642v2","created_at":"2026-06-29T01:14:28.092041+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.17642","created_at":"2026-06-29T01:14:28.092041+00:00"},{"alias_kind":"pith_short_12","alias_value":"MTZKXHAXDGJP","created_at":"2026-06-29T01:14:28.092041+00:00"},{"alias_kind":"pith_short_16","alias_value":"MTZKXHAXDGJPFJHY","created_at":"2026-06-29T01:14:28.092041+00:00"},{"alias_kind":"pith_short_8","alias_value":"MTZKXHAX","created_at":"2026-06-29T01:14:28.092041+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2607.02047","citing_title":"OpenSafeIntent: Evaluating Intent-Calibrated Safe Completion Across Dual-Use Prompt Sets","ref_index":31,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3","json":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3.json","graph_json":"https://pith.science/api/pith-number/MTZKXHAXDGJPFJHYOFLRESQMK3/graph.json","events_json":"https://pith.science/api/pith-number/MTZKXHAXDGJPFJHYOFLRESQMK3/events.json","paper":"https://pith.science/paper/MTZKXHAX"},"agent_actions":{"view_html":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3","download_json":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3.json","view_paper":"https://pith.science/paper/MTZKXHAX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.17642&json=true","fetch_graph":"https://pith.science/api/pith-number/MTZKXHAXDGJPFJHYOFLRESQMK3/graph.json","fetch_events":"https://pith.science/api/pith-number/MTZKXHAXDGJPFJHYOFLRESQMK3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3/action/storage_attestation","attest_author":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3/action/author_attestation","sign_citation":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3/action/citation_signature","submit_replication":"https://pith.science/pith/MTZKXHAXDGJPFJHYOFLRESQMK3/action/replication_record"}},"created_at":"2026-06-29T01:14:28.092041+00:00","updated_at":"2026-06-29T01:14:28.092041+00:00"}