{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:3ECVITF272CUM7L2G7OVQWO2ZX","short_pith_number":"pith:3ECVITF2","schema_version":"1.0","canonical_sha256":"d905544cbafe85467d7a37dd5859dacdd2296fb488f7b9d1ba10863f7bfb60f5","source":{"kind":"arxiv","id":"2412.15264","version":3},"attestation_state":"computed","paper":{"title":"ReXTrust: A Model for Fine-Grained Hallucination Detection in AI-Generated Radiology Reports","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Du Hyun Ro, Pranav Rajpurkar, Romain Hardy, Sung Eun Kim","submitted_at":"2024-12-17T02:07:33Z","abstract_excerpt":"The increasing adoption of AI-generated radiology reports necessitates robust methods for detecting hallucinations--false or unfounded statements that could impact patient care. We present ReXTrust, a novel framework for fine-grained hallucination detection in AI-generated radiology reports. Our approach leverages sequences of hidden states from large vision-language models to produce finding-level hallucination risk scores. We evaluate ReXTrust on a subset of the MIMIC-CXR dataset and demonstrate superior performance compared to existing approaches, achieving an AUROC of 0.8751 across all fin"},"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":"2412.15264","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-12-17T02:07:33Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"9b64c6ce9732c7444d83907aa10bc152837444357097d93f37212c0a4013fd8e","abstract_canon_sha256":"3cda125e38198e067516e8747edc3f4ee4ef240ba797d6af5f129d35eb136c13"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:07:47.099492Z","signature_b64":"R0pTjavUHOGi4feNddbnYgeLbOlMgR2QD7oKVaBtVuLQsYoyAkzkJjzhky7titglw6QLgPuDhkWmCknMYw0HDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d905544cbafe85467d7a37dd5859dacdd2296fb488f7b9d1ba10863f7bfb60f5","last_reissued_at":"2026-07-05T10:07:47.098974Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:07:47.098974Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReXTrust: A Model for Fine-Grained Hallucination Detection in AI-Generated Radiology Reports","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Du Hyun Ro, Pranav Rajpurkar, Romain Hardy, Sung Eun Kim","submitted_at":"2024-12-17T02:07:33Z","abstract_excerpt":"The increasing adoption of AI-generated radiology reports necessitates robust methods for detecting hallucinations--false or unfounded statements that could impact patient care. We present ReXTrust, a novel framework for fine-grained hallucination detection in AI-generated radiology reports. Our approach leverages sequences of hidden states from large vision-language models to produce finding-level hallucination risk scores. We evaluate ReXTrust on a subset of the MIMIC-CXR dataset and demonstrate superior performance compared to existing approaches, achieving an AUROC of 0.8751 across all fin"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2412.15264","kind":"arxiv","version":3},"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/2412.15264/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":"2412.15264","created_at":"2026-07-05T10:07:47.099032+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.15264v3","created_at":"2026-07-05T10:07:47.099032+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.15264","created_at":"2026-07-05T10:07:47.099032+00:00"},{"alias_kind":"pith_short_12","alias_value":"3ECVITF272CU","created_at":"2026-07-05T10:07:47.099032+00:00"},{"alias_kind":"pith_short_16","alias_value":"3ECVITF272CUM7L2","created_at":"2026-07-05T10:07:47.099032+00:00"},{"alias_kind":"pith_short_8","alias_value":"3ECVITF2","created_at":"2026-07-05T10:07:47.099032+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2508.07031","citing_title":"Trustworthy Medical Imaging with Large Language Models: A Study of Hallucinations Across Modalities","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX","json":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX.json","graph_json":"https://pith.science/api/pith-number/3ECVITF272CUM7L2G7OVQWO2ZX/graph.json","events_json":"https://pith.science/api/pith-number/3ECVITF272CUM7L2G7OVQWO2ZX/events.json","paper":"https://pith.science/paper/3ECVITF2"},"agent_actions":{"view_html":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX","download_json":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX.json","view_paper":"https://pith.science/paper/3ECVITF2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.15264&json=true","fetch_graph":"https://pith.science/api/pith-number/3ECVITF272CUM7L2G7OVQWO2ZX/graph.json","fetch_events":"https://pith.science/api/pith-number/3ECVITF272CUM7L2G7OVQWO2ZX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX/action/storage_attestation","attest_author":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX/action/author_attestation","sign_citation":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX/action/citation_signature","submit_replication":"https://pith.science/pith/3ECVITF272CUM7L2G7OVQWO2ZX/action/replication_record"}},"created_at":"2026-07-05T10:07:47.099032+00:00","updated_at":"2026-07-05T10:07:47.099032+00:00"}