{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:URUMUOIRFOKM4VOL7XJ2JIAQDY","short_pith_number":"pith:URUMUOIR","schema_version":"1.0","canonical_sha256":"a468ca39112b94ce55cbfdd3a4a0101e3377ceaebb776009b5896a9136e66c4c","source":{"kind":"arxiv","id":"2607.21065","version":1},"attestation_state":"computed","paper":{"title":"Do Pathology Vision-Language Models Truly See Pathology?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bo Li, Chengyang Zhang, Chenxun Deng, Hong Bu, Jiaming Yang, Jiancheng Lv, Jie Chen, Mengran Li, Wei Ju, Wenchuan Zhang, Xinyu Liu, Yang Zhang, Yuhao Yi","submitted_at":"2026-07-23T08:59:11Z","abstract_excerpt":"Pathology vision-language models (VLMs) have recently progressed rapidly and are commonly evaluated by answer accuracy on pathology VQA benchmarks. However, we dig into current evaluations and identify three overlooked issues: 1) Visual evidence is not always necessary. For instance, Gemini-3-Pro achieves 53.5% average accuracy across 5 VQA benchmarks without any visual input. 2) Domain training can improve accuracy without proportional gains in visual binding. Compared with Qwen2.5-VL-7B, Patho-R1-7B exhibits a 5.8-point lower multimodal gain and a 3.7-point lower attention IoU. 3) Entity-lev"},"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":"2607.21065","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-07-23T08:59:11Z","cross_cats_sorted":[],"title_canon_sha256":"255489ac4baa5d322417dc335ea2bb0304d2d38d748c0c9c89f25085e4c32809","abstract_canon_sha256":"c833e5fc316e7ba679fcf3020d35791ab1684bfc08db37bcdc28075522a10063"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-24T01:23:50.020660Z","signature_b64":"Brkbch9kApoMk1GicNN1yywbDUrdoNvZavXA2ahSuucaJhyFNcMORJe/JMqyAgy2mEYMirA1AEII/i27tELHDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a468ca39112b94ce55cbfdd3a4a0101e3377ceaebb776009b5896a9136e66c4c","last_reissued_at":"2026-07-24T01:23:50.019761Z","signature_status":"signed_v1","first_computed_at":"2026-07-24T01:23:50.019761Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Do Pathology Vision-Language Models Truly See Pathology?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Bo Li, Chengyang Zhang, Chenxun Deng, Hong Bu, Jiaming Yang, Jiancheng Lv, Jie Chen, Mengran Li, Wei Ju, Wenchuan Zhang, Xinyu Liu, Yang Zhang, Yuhao Yi","submitted_at":"2026-07-23T08:59:11Z","abstract_excerpt":"Pathology vision-language models (VLMs) have recently progressed rapidly and are commonly evaluated by answer accuracy on pathology VQA benchmarks. However, we dig into current evaluations and identify three overlooked issues: 1) Visual evidence is not always necessary. For instance, Gemini-3-Pro achieves 53.5% average accuracy across 5 VQA benchmarks without any visual input. 2) Domain training can improve accuracy without proportional gains in visual binding. Compared with Qwen2.5-VL-7B, Patho-R1-7B exhibits a 5.8-point lower multimodal gain and a 3.7-point lower attention IoU. 3) Entity-lev"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.21065","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/2607.21065/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":"2607.21065","created_at":"2026-07-24T01:23:50.020218+00:00"},{"alias_kind":"arxiv_version","alias_value":"2607.21065v1","created_at":"2026-07-24T01:23:50.020218+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.21065","created_at":"2026-07-24T01:23:50.020218+00:00"},{"alias_kind":"pith_short_12","alias_value":"URUMUOIRFOKM","created_at":"2026-07-24T01:23:50.020218+00:00"},{"alias_kind":"pith_short_16","alias_value":"URUMUOIRFOKM4VOL","created_at":"2026-07-24T01:23:50.020218+00:00"},{"alias_kind":"pith_short_8","alias_value":"URUMUOIR","created_at":"2026-07-24T01:23:50.020218+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/URUMUOIRFOKM4VOL7XJ2JIAQDY","json":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY.json","graph_json":"https://pith.science/api/pith-number/URUMUOIRFOKM4VOL7XJ2JIAQDY/graph.json","events_json":"https://pith.science/api/pith-number/URUMUOIRFOKM4VOL7XJ2JIAQDY/events.json","paper":"https://pith.science/paper/URUMUOIR"},"agent_actions":{"view_html":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY","download_json":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY.json","view_paper":"https://pith.science/paper/URUMUOIR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2607.21065&json=true","fetch_graph":"https://pith.science/api/pith-number/URUMUOIRFOKM4VOL7XJ2JIAQDY/graph.json","fetch_events":"https://pith.science/api/pith-number/URUMUOIRFOKM4VOL7XJ2JIAQDY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY/action/storage_attestation","attest_author":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY/action/author_attestation","sign_citation":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY/action/citation_signature","submit_replication":"https://pith.science/pith/URUMUOIRFOKM4VOL7XJ2JIAQDY/action/replication_record"}},"created_at":"2026-07-24T01:23:50.020218+00:00","updated_at":"2026-07-24T01:23:50.020218+00:00"}