{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:QIEWJ6WA7IRTM5RQHUVTPVGT6W","short_pith_number":"pith:QIEWJ6WA","schema_version":"1.0","canonical_sha256":"820964fac0fa233676303d2b37d4d3f58f205f40ef9aff820cdeb364ae1473b7","source":{"kind":"arxiv","id":"2502.08079","version":3},"attestation_state":"computed","paper":{"title":"MAA: Meticulous Adversarial Attack against Vision-Language Pre-trained Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Guangdong Bai, Peng-fei Zhang, Zi Huang","submitted_at":"2025-02-12T02:53:27Z","abstract_excerpt":"Current adversarial attacks for evaluating the robustness of vision-language pre-trained (VLP) models in multi-modal tasks suffer from limited transferability, where attacks crafted for a specific model often struggle to generalize effectively across different models, limiting their utility in assessing robustness more broadly. This is mainly attributed to the over-reliance on model-specific features and regions, particularly in the image modality. In this paper, we propose an elegant yet highly effective method termed Meticulous Adversarial Attack (MAA) to fully exploit model-independent char"},"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":"2502.08079","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2025-02-12T02:53:27Z","cross_cats_sorted":[],"title_canon_sha256":"5f73e96d5db3249a88dd4faa60bae1243bfc0e16c51054a45325abecc386e654","abstract_canon_sha256":"d2c8e2cb7ae89fb1868a79fcbdbebd6cfcc1a4102273627c8bbc03a30a7d9640"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:22:13.151089Z","signature_b64":"lf/zS1IapfHqxmkxsD0uMQc0NDz7rdFFbPgK1e3jesJ24U9fE+vEwTeaSy8fFkAxfLLCZOragg2zFDrWmKGmDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"820964fac0fa233676303d2b37d4d3f58f205f40ef9aff820cdeb364ae1473b7","last_reissued_at":"2026-07-05T10:22:13.150476Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:22:13.150476Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MAA: Meticulous Adversarial Attack against Vision-Language Pre-trained Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Guangdong Bai, Peng-fei Zhang, Zi Huang","submitted_at":"2025-02-12T02:53:27Z","abstract_excerpt":"Current adversarial attacks for evaluating the robustness of vision-language pre-trained (VLP) models in multi-modal tasks suffer from limited transferability, where attacks crafted for a specific model often struggle to generalize effectively across different models, limiting their utility in assessing robustness more broadly. This is mainly attributed to the over-reliance on model-specific features and regions, particularly in the image modality. In this paper, we propose an elegant yet highly effective method termed Meticulous Adversarial Attack (MAA) to fully exploit model-independent char"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.08079","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/2502.08079/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":"2502.08079","created_at":"2026-07-05T10:22:13.150560+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.08079v3","created_at":"2026-07-05T10:22:13.150560+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.08079","created_at":"2026-07-05T10:22:13.150560+00:00"},{"alias_kind":"pith_short_12","alias_value":"QIEWJ6WA7IRT","created_at":"2026-07-05T10:22:13.150560+00:00"},{"alias_kind":"pith_short_16","alias_value":"QIEWJ6WA7IRTM5RQ","created_at":"2026-07-05T10:22:13.150560+00:00"},{"alias_kind":"pith_short_8","alias_value":"QIEWJ6WA","created_at":"2026-07-05T10:22:13.150560+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26199","citing_title":"MIRAGE: Protecting against Malicious Image Editing via False Moderation","ref_index":81,"is_internal_anchor":false},{"citing_arxiv_id":"2606.26199","citing_title":"MIRAGE: Protecting against Malicious Image Editing via False Moderation","ref_index":81,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W","json":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W.json","graph_json":"https://pith.science/api/pith-number/QIEWJ6WA7IRTM5RQHUVTPVGT6W/graph.json","events_json":"https://pith.science/api/pith-number/QIEWJ6WA7IRTM5RQHUVTPVGT6W/events.json","paper":"https://pith.science/paper/QIEWJ6WA"},"agent_actions":{"view_html":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W","download_json":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W.json","view_paper":"https://pith.science/paper/QIEWJ6WA","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.08079&json=true","fetch_graph":"https://pith.science/api/pith-number/QIEWJ6WA7IRTM5RQHUVTPVGT6W/graph.json","fetch_events":"https://pith.science/api/pith-number/QIEWJ6WA7IRTM5RQHUVTPVGT6W/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W/action/storage_attestation","attest_author":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W/action/author_attestation","sign_citation":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W/action/citation_signature","submit_replication":"https://pith.science/pith/QIEWJ6WA7IRTM5RQHUVTPVGT6W/action/replication_record"}},"created_at":"2026-07-05T10:22:13.150560+00:00","updated_at":"2026-07-05T10:22:13.150560+00:00"}