{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:SAWPJNJM6TEPYZ7PZUNA2NV7RB","short_pith_number":"pith:SAWPJNJM","schema_version":"1.0","canonical_sha256":"902cf4b52cf4c8fc67efcd1a0d36bf885a66ed8cb38b45c8c65111e381f99d89","source":{"kind":"arxiv","id":"2606.28900","version":1},"attestation_state":"computed","paper":{"title":"MedEvoEval: Evaluating Continual Evolution of Doctor Agents through Simulated Clinical Episodes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Hui Zhang","submitted_at":"2026-06-27T13:14:16Z","abstract_excerpt":"Doctor agents are moving beyond single-turn answer generation toward evolving clinical decision systems. Within an outpatient episode, they acquire evidence, use examination and consultation resources, and decide when to finalize a diagnosis and management plan. Across episodes, their behavior may change through memory, retrieval, reflection, or other update mechanisms. Current evaluations only partially cover this setting. Fixed-input medical QA benchmarks score final answers from complete inputs, whereas many interactive benchmarks still focus on individual encounters or fixed runs, providin"},"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":"2606.28900","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-06-27T13:14:16Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"2893526e37d8bcdf079c52864dffbb6a33ce1b9b21e5a61b46efb7fc4920b04d","abstract_canon_sha256":"d05e9ddc1ce95b160b5ee67b220396657b88b0f099f1778a07851964094b6755"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-30T01:17:44.347960Z","signature_b64":"cg70FxO5R1dEp5T9RYDqb6HPud0KBB+9dRm02hyM4ey8Z2OZ83UL01UplmcfBWRCaYRFFLApQ3Olf446YMWEDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"902cf4b52cf4c8fc67efcd1a0d36bf885a66ed8cb38b45c8c65111e381f99d89","last_reissued_at":"2026-06-30T01:17:44.347236Z","signature_status":"signed_v1","first_computed_at":"2026-06-30T01:17:44.347236Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MedEvoEval: Evaluating Continual Evolution of Doctor Agents through Simulated Clinical Episodes","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Hui Zhang","submitted_at":"2026-06-27T13:14:16Z","abstract_excerpt":"Doctor agents are moving beyond single-turn answer generation toward evolving clinical decision systems. Within an outpatient episode, they acquire evidence, use examination and consultation resources, and decide when to finalize a diagnosis and management plan. Across episodes, their behavior may change through memory, retrieval, reflection, or other update mechanisms. Current evaluations only partially cover this setting. Fixed-input medical QA benchmarks score final answers from complete inputs, whereas many interactive benchmarks still focus on individual encounters or fixed runs, providin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.28900","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/2606.28900/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":"2606.28900","created_at":"2026-06-30T01:17:44.347370+00:00"},{"alias_kind":"arxiv_version","alias_value":"2606.28900v1","created_at":"2026-06-30T01:17:44.347370+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.28900","created_at":"2026-06-30T01:17:44.347370+00:00"},{"alias_kind":"pith_short_12","alias_value":"SAWPJNJM6TEP","created_at":"2026-06-30T01:17:44.347370+00:00"},{"alias_kind":"pith_short_16","alias_value":"SAWPJNJM6TEPYZ7P","created_at":"2026-06-30T01:17:44.347370+00:00"},{"alias_kind":"pith_short_8","alias_value":"SAWPJNJM","created_at":"2026-06-30T01:17:44.347370+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/SAWPJNJM6TEPYZ7PZUNA2NV7RB","json":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB.json","graph_json":"https://pith.science/api/pith-number/SAWPJNJM6TEPYZ7PZUNA2NV7RB/graph.json","events_json":"https://pith.science/api/pith-number/SAWPJNJM6TEPYZ7PZUNA2NV7RB/events.json","paper":"https://pith.science/paper/SAWPJNJM"},"agent_actions":{"view_html":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB","download_json":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB.json","view_paper":"https://pith.science/paper/SAWPJNJM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2606.28900&json=true","fetch_graph":"https://pith.science/api/pith-number/SAWPJNJM6TEPYZ7PZUNA2NV7RB/graph.json","fetch_events":"https://pith.science/api/pith-number/SAWPJNJM6TEPYZ7PZUNA2NV7RB/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB/action/storage_attestation","attest_author":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB/action/author_attestation","sign_citation":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB/action/citation_signature","submit_replication":"https://pith.science/pith/SAWPJNJM6TEPYZ7PZUNA2NV7RB/action/replication_record"}},"created_at":"2026-06-30T01:17:44.347370+00:00","updated_at":"2026-06-30T01:17:44.347370+00:00"}