{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:QU7AWHWVZPO5EW2K3DU5NWPAWM","short_pith_number":"pith:QU7AWHWV","schema_version":"1.0","canonical_sha256":"853e0b1ed5cbddd25b4ad8e9d6d9e0b3042c443b5667ab98b7038aabaef5f034","source":{"kind":"arxiv","id":"2608.04885","version":1},"attestation_state":"computed","paper":{"title":"Evaluating the Diagnostic Robustness of Vision-Language Models Under Visual and Textual Perturbations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ali Khoramfar, Alireza Mohamadian, Heshaam Faili, Mohammad Javad Dousti","submitted_at":"2026-08-05T14:09:55Z","abstract_excerpt":"Standard accuracy metrics for VLMs often mask significant reliability failures in sensitive domains. In this work, we utilize a histopathology-validated brain MRI dataset to systematically assess the diagnostic robustness of four VLM families under evidence-preserving perturbations. By reordering anatomical slices and swapping target label positions, we evaluate whether models maintain consistent predictions when clinical evidence remains invariant. Our results reveal significant vulnerabilities in presentation-order stability, with models exhibiting prediction flips in up to 48.9% of cases un"},"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":"2608.04885","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2026-08-05T14:09:55Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"0c5a500c87ed7962d54bda389ff60e29fe23d4c88746739c4ad115cefd5235a2","abstract_canon_sha256":"957b5bcaa2413bde8720e2436a6eb52fab7af7cfc93504b97b2059b3661f4fbc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-06T01:47:27.445083Z","signature_b64":"/xP6qjksoWumbLPv5TkN4+OIPeRMSwb5Q8TQW0b8w8zupwldoDGfc66v1bjY7cqFwo1UWRmsFVoQKQxRKZdMDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"853e0b1ed5cbddd25b4ad8e9d6d9e0b3042c443b5667ab98b7038aabaef5f034","last_reissued_at":"2026-08-06T01:47:27.443652Z","signature_status":"signed_v1","first_computed_at":"2026-08-06T01:47:27.443652Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Evaluating the Diagnostic Robustness of Vision-Language Models Under Visual and Textual Perturbations","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Ali Khoramfar, Alireza Mohamadian, Heshaam Faili, Mohammad Javad Dousti","submitted_at":"2026-08-05T14:09:55Z","abstract_excerpt":"Standard accuracy metrics for VLMs often mask significant reliability failures in sensitive domains. In this work, we utilize a histopathology-validated brain MRI dataset to systematically assess the diagnostic robustness of four VLM families under evidence-preserving perturbations. By reordering anatomical slices and swapping target label positions, we evaluate whether models maintain consistent predictions when clinical evidence remains invariant. Our results reveal significant vulnerabilities in presentation-order stability, with models exhibiting prediction flips in up to 48.9% of cases un"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.04885","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/2608.04885/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":"2608.04885","created_at":"2026-08-06T01:47:27.445500+00:00"},{"alias_kind":"arxiv_version","alias_value":"2608.04885v1","created_at":"2026-08-06T01:47:27.445500+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.04885","created_at":"2026-08-06T01:47:27.445500+00:00"},{"alias_kind":"pith_short_12","alias_value":"QU7AWHWVZPO5","created_at":"2026-08-06T01:47:27.445500+00:00"},{"alias_kind":"pith_short_16","alias_value":"QU7AWHWVZPO5EW2K","created_at":"2026-08-06T01:47:27.445500+00:00"},{"alias_kind":"pith_short_8","alias_value":"QU7AWHWV","created_at":"2026-08-06T01:47:27.445500+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/QU7AWHWVZPO5EW2K3DU5NWPAWM","json":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM.json","graph_json":"https://pith.science/api/pith-number/QU7AWHWVZPO5EW2K3DU5NWPAWM/graph.json","events_json":"https://pith.science/api/pith-number/QU7AWHWVZPO5EW2K3DU5NWPAWM/events.json","paper":"https://pith.science/paper/QU7AWHWV"},"agent_actions":{"view_html":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM","download_json":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM.json","view_paper":"https://pith.science/paper/QU7AWHWV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2608.04885&json=true","fetch_graph":"https://pith.science/api/pith-number/QU7AWHWVZPO5EW2K3DU5NWPAWM/graph.json","fetch_events":"https://pith.science/api/pith-number/QU7AWHWVZPO5EW2K3DU5NWPAWM/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM/action/storage_attestation","attest_author":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM/action/author_attestation","sign_citation":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM/action/citation_signature","submit_replication":"https://pith.science/pith/QU7AWHWVZPO5EW2K3DU5NWPAWM/action/replication_record"}},"created_at":"2026-08-06T01:47:27.445500+00:00","updated_at":"2026-08-06T01:47:27.445500+00:00"}