{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:GFNULGWHOXNG4N7SKKLZKRQTN3","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"99c92429c6898478d4948c0e035da1e3e707231ec0f6028a7dbb3933d9413592","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-24T06:58:57Z","title_canon_sha256":"0ce09ff269858414f59e1c274d33cac47216e04dcd76758fafdaeea2a8c7133b"},"schema_version":"1.0","source":{"id":"2605.24902","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.24902","created_at":"2026-05-26T01:04:04Z"},{"alias_kind":"arxiv_version","alias_value":"2605.24902v1","created_at":"2026-05-26T01:04:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.24902","created_at":"2026-05-26T01:04:04Z"},{"alias_kind":"pith_short_12","alias_value":"GFNULGWHOXNG","created_at":"2026-05-26T01:04:04Z"},{"alias_kind":"pith_short_16","alias_value":"GFNULGWHOXNG4N7S","created_at":"2026-05-26T01:04:04Z"},{"alias_kind":"pith_short_8","alias_value":"GFNULGWH","created_at":"2026-05-26T01:04:04Z"}],"graph_snapshots":[{"event_id":"sha256:d1d60c82e868ad3b0327073900eb46a7507fd08078702c9eef2215336ea4232c","target":"graph","created_at":"2026-05-26T01:04:04Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2605.24902/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reasoning-enabled LLMs perform strongly on medical reasoning benchmarks, but it remains unclear whether these gains transfer to structured clinical documentation; we investigate this question using SOAP note generation from clinical dialogue in a source-aware benchmark spanning OMI Health, ACI-Bench, and PriMock57. We evaluate GPT-5.4, DeepSeek-V4-Flash, and Gemma-4-E4B in a controlled 2x2 design that independently toggles provider-native reasoning and same-source retrieval-augmented generation (RAG). Outputs are assessed using seven automatic metrics alongside two reference-aware LLM judges. ","authors_text":"Faizan Faisal","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-24T06:58:57Z","title":"When Reasoning Hurts: Source-Aware Evaluation of Frontier LLMs for Clinical SOAP Note Generation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.24902","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:887a2f0a1507ec6bb75db12c5cbc8d76e12b0399d6e3c8e319159a3df2b0c0b7","target":"record","created_at":"2026-05-26T01:04:04Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"99c92429c6898478d4948c0e035da1e3e707231ec0f6028a7dbb3933d9413592","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-05-24T06:58:57Z","title_canon_sha256":"0ce09ff269858414f59e1c274d33cac47216e04dcd76758fafdaeea2a8c7133b"},"schema_version":"1.0","source":{"id":"2605.24902","kind":"arxiv","version":1}},"canonical_sha256":"315b459ac775da6e37f252979546136ed545b1db87b10651da48c1414183d9da","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"315b459ac775da6e37f252979546136ed545b1db87b10651da48c1414183d9da","first_computed_at":"2026-05-26T01:04:04.588098Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-26T01:04:04.588098Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"OPcsXFEf5hZSuhuQO+fSpqqfKHPt4hmTNJV0CwecQa7N5/UvQYUY5i4UgdLEprgVv7+tp1moOXktZHPsoJkLAg==","signature_status":"signed_v1","signed_at":"2026-05-26T01:04:04.588828Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.24902","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:887a2f0a1507ec6bb75db12c5cbc8d76e12b0399d6e3c8e319159a3df2b0c0b7","sha256:d1d60c82e868ad3b0327073900eb46a7507fd08078702c9eef2215336ea4232c"],"state_sha256":"2e3658238c5dbaa8696eb2ddd3b9df80d037153d5579c8a678a10e6970450e76"}