{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:XVPWY4TI2IFKGNKNTOBL3YTXO2","short_pith_number":"pith:XVPWY4TI","schema_version":"1.0","canonical_sha256":"bd5f6c7268d20aa3354d9b82bde27776b11baac658dabc8f66d716c9113b22f3","source":{"kind":"arxiv","id":"2607.18999","version":1},"attestation_state":"computed","paper":{"title":"MedDDC-Eval: Diagnosis-Decoupled Evaluation of Multi-Turn Medical Consultation Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guofeng Zhang, Huaiyi Fang, Jianwei Lv, Jinyao Liu, Junfeng Wang, Lening An, Xunxu Duan, Yizeng Quan, Yu Ouyang","submitted_at":"2026-07-21T11:32:41Z","abstract_excerpt":"Multi-turn medical consultation agents must decide what to ask, adapt to patient responses, and determine when the collected evidence is sufficient. However, coupled evaluation conflates the quality of the policy-elicited history with policy-specific terminal diagnosis generation: strong generation can compensate for a thin history, while weaker generation can obscure a rich one. We introduce MedDDC-Eval, a diagnosis-decoupled testbed that treats elicited history as the comparison object and holds the history-to-diagnosis mapping constant through a shared frozen reader. Across two held-out sou"},"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":"2607.18999","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-07-21T11:32:41Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"3c36c105511bf77871b270570dce3172bf807e6572b35f12bb7a99d9ad774431","abstract_canon_sha256":"96c369b65b085e2ebb12123b477b41374be33198a40301c320ede1d270182409"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-22T01:24:03.838404Z","signature_b64":"h+oj9E8DSVMdHIa6RhzLl6lFKHjQRH/JySGOxDxDuDYlcPjI0d6VoyyUgzePBuRJCp2SfWNQ49+z/XR/Sa8BAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bd5f6c7268d20aa3354d9b82bde27776b11baac658dabc8f66d716c9113b22f3","last_reissued_at":"2026-07-22T01:24:03.837662Z","signature_status":"signed_v1","first_computed_at":"2026-07-22T01:24:03.837662Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MedDDC-Eval: Diagnosis-Decoupled Evaluation of Multi-Turn Medical Consultation Agents","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Guofeng Zhang, Huaiyi Fang, Jianwei Lv, Jinyao Liu, Junfeng Wang, Lening An, Xunxu Duan, Yizeng Quan, Yu Ouyang","submitted_at":"2026-07-21T11:32:41Z","abstract_excerpt":"Multi-turn medical consultation agents must decide what to ask, adapt to patient responses, and determine when the collected evidence is sufficient. However, coupled evaluation conflates the quality of the policy-elicited history with policy-specific terminal diagnosis generation: strong generation can compensate for a thin history, while weaker generation can obscure a rich one. We introduce MedDDC-Eval, a diagnosis-decoupled testbed that treats elicited history as the comparison object and holds the history-to-diagnosis mapping constant through a shared frozen reader. Across two held-out sou"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.18999","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/2607.18999/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":"2607.18999","created_at":"2026-07-22T01:24:03.838066+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.18999v1","created_at":"2026-07-22T01:24:03.838066+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.18999","created_at":"2026-07-22T01:24:03.838066+00:00"},{"alias_kind":"pith_short_12","alias_value":"XVPWY4TI2IFK","created_at":"2026-07-22T01:24:03.838066+00:00"},{"alias_kind":"pith_short_16","alias_value":"XVPWY4TI2IFKGNKN","created_at":"2026-07-22T01:24:03.838066+00:00"},{"alias_kind":"pith_short_8","alias_value":"XVPWY4TI","created_at":"2026-07-22T01:24:03.838066+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/XVPWY4TI2IFKGNKNTOBL3YTXO2","json":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2.json","graph_json":"https://pith.science/api/pith-number/XVPWY4TI2IFKGNKNTOBL3YTXO2/graph.json","events_json":"https://pith.science/api/pith-number/XVPWY4TI2IFKGNKNTOBL3YTXO2/events.json","paper":"https://pith.science/paper/XVPWY4TI"},"agent_actions":{"view_html":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2","download_json":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2.json","view_paper":"https://pith.science/paper/XVPWY4TI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.18999&json=true","fetch_graph":"https://pith.science/api/pith-number/XVPWY4TI2IFKGNKNTOBL3YTXO2/graph.json","fetch_events":"https://pith.science/api/pith-number/XVPWY4TI2IFKGNKNTOBL3YTXO2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2/action/storage_attestation","attest_author":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2/action/author_attestation","sign_citation":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2/action/citation_signature","submit_replication":"https://pith.science/pith/XVPWY4TI2IFKGNKNTOBL3YTXO2/action/replication_record"}},"created_at":"2026-07-22T01:24:03.838066+00:00","updated_at":"2026-07-22T01:24:03.838066+00:00"}