{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:EE32JH4DGZZEN26ZBEF4PTGGQI","short_pith_number":"pith:EE32JH4D","schema_version":"1.0","canonical_sha256":"2137a49f83367246ebd9090bc7ccc6823cf9b6a35c22e5f88bbf2e2843b6cc76","source":{"kind":"arxiv","id":"2607.20462","version":1},"attestation_state":"computed","paper":{"title":"Marking the Wrong Symptoms: Evaluating LLM Watermarks in Medical Texts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Martin Vechev, Melanie Rieff, Robin Staab, Stefan Hegselmann, Thibaud Gloaguen","submitted_at":"2026-05-16T08:16:08Z","abstract_excerpt":"Large language models (LLMs) are increasingly integrated into clinical workflows, stressing the need for reliable traceability of model-generated output with watermarking. Yet, most watermarks are evaluated on general-purpose benchmarks, leaving domains like medicine, where small token-level perturbations can result in significant semantic changes, underexplored. In this work, we present the first rigorous study of how LLM watermarks affect medical performance, benchmarking 5 watermarking schemes across 11 LLMs and 7 VLMs on various tasks spanning unimodal and multimodal clinical reasoning. Im"},"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":"2607.20462","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-05-16T08:16:08Z","cross_cats_sorted":[],"title_canon_sha256":"762ca303e01e62c708d1cda5134bd04b0c552048eef0648f082e62394343ae7f","abstract_canon_sha256":"a5c605e9d3605ab3fd3ee0d00e43481f97ff58dff35ef3b59c4705c9386138dc"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-24T00:23:16.851940Z","signature_b64":"OpuDBl9Wbcr12E+55TptFWzMzsd0l8M3/5LlbiLP6tpjSWaP0cZjJSDfP70fHNwQIrlZrokZtzn6IGD/e+RZBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2137a49f83367246ebd9090bc7ccc6823cf9b6a35c22e5f88bbf2e2843b6cc76","last_reissued_at":"2026-07-24T00:23:16.851076Z","signature_status":"signed_v1","first_computed_at":"2026-07-24T00:23:16.851076Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Marking the Wrong Symptoms: Evaluating LLM Watermarks in Medical Texts","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Martin Vechev, Melanie Rieff, Robin Staab, Stefan Hegselmann, Thibaud Gloaguen","submitted_at":"2026-05-16T08:16:08Z","abstract_excerpt":"Large language models (LLMs) are increasingly integrated into clinical workflows, stressing the need for reliable traceability of model-generated output with watermarking. Yet, most watermarks are evaluated on general-purpose benchmarks, leaving domains like medicine, where small token-level perturbations can result in significant semantic changes, underexplored. In this work, we present the first rigorous study of how LLM watermarks affect medical performance, benchmarking 5 watermarking schemes across 11 LLMs and 7 VLMs on various tasks spanning unimodal and multimodal clinical reasoning. Im"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.20462","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/2607.20462/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":"2607.20462","created_at":"2026-07-24T00:23:16.851522+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.20462v1","created_at":"2026-07-24T00:23:16.851522+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.20462","created_at":"2026-07-24T00:23:16.851522+00:00"},{"alias_kind":"pith_short_12","alias_value":"EE32JH4DGZZE","created_at":"2026-07-24T00:23:16.851522+00:00"},{"alias_kind":"pith_short_16","alias_value":"EE32JH4DGZZEN26Z","created_at":"2026-07-24T00:23:16.851522+00:00"},{"alias_kind":"pith_short_8","alias_value":"EE32JH4D","created_at":"2026-07-24T00:23:16.851522+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/EE32JH4DGZZEN26ZBEF4PTGGQI","json":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI.json","graph_json":"https://pith.science/api/pith-number/EE32JH4DGZZEN26ZBEF4PTGGQI/graph.json","events_json":"https://pith.science/api/pith-number/EE32JH4DGZZEN26ZBEF4PTGGQI/events.json","paper":"https://pith.science/paper/EE32JH4D"},"agent_actions":{"view_html":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI","download_json":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI.json","view_paper":"https://pith.science/paper/EE32JH4D","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.20462&json=true","fetch_graph":"https://pith.science/api/pith-number/EE32JH4DGZZEN26ZBEF4PTGGQI/graph.json","fetch_events":"https://pith.science/api/pith-number/EE32JH4DGZZEN26ZBEF4PTGGQI/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI/action/timestamp_anchor","attest_storage":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI/action/storage_attestation","attest_author":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI/action/author_attestation","sign_citation":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI/action/citation_signature","submit_replication":"https://pith.science/pith/EE32JH4DGZZEN26ZBEF4PTGGQI/action/replication_record"}},"created_at":"2026-07-24T00:23:16.851522+00:00","updated_at":"2026-07-24T00:23:16.851522+00:00"}