{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UV7MQPAYYQEXZPQL7DSGUC2QXV","short_pith_number":"pith:UV7MQPAY","schema_version":"1.0","canonical_sha256":"a57ec83c18c4097cbe0bf8e46a0b50bd63e0dd98832ba5e3162b1ba014768176","source":{"kind":"arxiv","id":"2601.12263","version":2},"attestation_state":"computed","paper":{"title":"Multimodal Generative Engine Optimization: Rank Manipulation for Vision-Language Model Rankers","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Chenxiao Yu, Haoyan Xu, Xiyang Hu, Yixuan Du, Yue Zhao, Ziyi Wang","submitted_at":"2026-01-18T04:58:28Z","abstract_excerpt":"Vision-Language Models (VLMs) integrate visual and textual knowledge into unified representations that increasingly underpin modern retrieval and recommendation systems. However, it remains unclear how reliably these models utilize their cross-modal knowledge when ranking multimodal items, and whether their knowledge grounding can be subverted. In this paper, we expose a fundamental vulnerability in how VLMs apply multimodal knowledge for product ranking: through Multimodal Generative Engine Optimization (MGEO), we show that an adversary can manipulate a VLM's ranking decisions by jointly craf"},"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":"2601.12263","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2026-01-18T04:58:28Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"b1e03d0e564dd2e20f20e147b5ded9315b53a76cd10fd329d7cfc86a5ea6b7d8","abstract_canon_sha256":"1beacebb9f50f08a5856fe6718a504e34522a1b077fd75b01d0c6ca3bb564568"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-09T02:07:17.571758Z","signature_b64":"qs4f1eMNA+cbtT90uDxjC27eQ93bhbXMzo1RRUf3cPxLLq5E/hhL893b4qiXl2+yEexSApOmDORQm3g0Y39GAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a57ec83c18c4097cbe0bf8e46a0b50bd63e0dd98832ba5e3162b1ba014768176","last_reissued_at":"2026-06-09T02:07:17.570745Z","signature_status":"signed_v1","first_computed_at":"2026-06-09T02:07:17.570745Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Multimodal Generative Engine Optimization: Rank Manipulation for Vision-Language Model Rankers","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Chenxiao Yu, Haoyan Xu, Xiyang Hu, Yixuan Du, Yue Zhao, Ziyi Wang","submitted_at":"2026-01-18T04:58:28Z","abstract_excerpt":"Vision-Language Models (VLMs) integrate visual and textual knowledge into unified representations that increasingly underpin modern retrieval and recommendation systems. However, it remains unclear how reliably these models utilize their cross-modal knowledge when ranking multimodal items, and whether their knowledge grounding can be subverted. In this paper, we expose a fundamental vulnerability in how VLMs apply multimodal knowledge for product ranking: through Multimodal Generative Engine Optimization (MGEO), we show that an adversary can manipulate a VLM's ranking decisions by jointly craf"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.12263","kind":"arxiv","version":2},"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/2601.12263/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":"2601.12263","created_at":"2026-06-09T02:07:17.570883+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.12263v2","created_at":"2026-06-09T02:07:17.570883+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.12263","created_at":"2026-06-09T02:07:17.570883+00:00"},{"alias_kind":"pith_short_12","alias_value":"UV7MQPAYYQEX","created_at":"2026-06-09T02:07:17.570883+00:00"},{"alias_kind":"pith_short_16","alias_value":"UV7MQPAYYQEXZPQL","created_at":"2026-06-09T02:07:17.570883+00:00"},{"alias_kind":"pith_short_8","alias_value":"UV7MQPAY","created_at":"2026-06-09T02:07:17.570883+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/UV7MQPAYYQEXZPQL7DSGUC2QXV","json":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV.json","graph_json":"https://pith.science/api/pith-number/UV7MQPAYYQEXZPQL7DSGUC2QXV/graph.json","events_json":"https://pith.science/api/pith-number/UV7MQPAYYQEXZPQL7DSGUC2QXV/events.json","paper":"https://pith.science/paper/UV7MQPAY"},"agent_actions":{"view_html":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV","download_json":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV.json","view_paper":"https://pith.science/paper/UV7MQPAY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.12263&json=true","fetch_graph":"https://pith.science/api/pith-number/UV7MQPAYYQEXZPQL7DSGUC2QXV/graph.json","fetch_events":"https://pith.science/api/pith-number/UV7MQPAYYQEXZPQL7DSGUC2QXV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV/action/storage_attestation","attest_author":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV/action/author_attestation","sign_citation":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV/action/citation_signature","submit_replication":"https://pith.science/pith/UV7MQPAYYQEXZPQL7DSGUC2QXV/action/replication_record"}},"created_at":"2026-06-09T02:07:17.570883+00:00","updated_at":"2026-06-09T02:07:17.570883+00:00"}