{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:6GCICKCRNHRTDBWJHOR6EKAPTK","short_pith_number":"pith:6GCICKCR","schema_version":"1.0","canonical_sha256":"f18481285169e33186c93ba3e2280f9a82e8cb7d40281843daa1d799d0d5bf91","source":{"kind":"arxiv","id":"2406.09250","version":3},"attestation_state":"computed","paper":{"title":"MirrorCheck: Efficient Adversarial Defense for Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Ivan Laptev, Karthik Nandakumar, Klea Ziu, Martin Tak\\'a\\v{c}, Nikita Durasov, Pascal Fua, Samar Fares, Toluwani Aremu","submitted_at":"2024-06-13T15:55:04Z","abstract_excerpt":"Vision-Language Models (VLMs) are increasingly susceptible to sophisticated adversarial attacks, including adaptive strategies specifically designed to bypass existing defenses. To address this vulnerability, we propose MirrorCheck, a robust and model-agnostic detection framework that operates effectively in both unimodal and multimodal settings. MirrorCheck leverages Text-to-Image (T2I) models to regenerate visual content from captions produced by the target model and assesses semantic consistency by comparing feature-space embeddings between the original and synthesized images. To enhance ro"},"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":"2406.09250","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2024-06-13T15:55:04Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"cba908ab5f0303a722ecd91e10d94d9c66b9179ca64b2e2c122ccc28b76a67d1","abstract_canon_sha256":"b96680902b50cf69427c6688dd4ccc051a77152eed841e5680d7d6f92c9ddb6f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-25T02:01:00.959393Z","signature_b64":"LwYu3zqBSrqKcp3Q+4vG6y3CxLaJ7eFxHnlDl7pBpuaqz7J5yv5L161V64fkbr9/pF48PFDUsXTtyBhdpwORBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f18481285169e33186c93ba3e2280f9a82e8cb7d40281843daa1d799d0d5bf91","last_reissued_at":"2026-05-25T02:01:00.958278Z","signature_status":"signed_v1","first_computed_at":"2026-05-25T02:01:00.958278Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"MirrorCheck: Efficient Adversarial Defense for Vision-Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CV","authors_text":"Ivan Laptev, Karthik Nandakumar, Klea Ziu, Martin Tak\\'a\\v{c}, Nikita Durasov, Pascal Fua, Samar Fares, Toluwani Aremu","submitted_at":"2024-06-13T15:55:04Z","abstract_excerpt":"Vision-Language Models (VLMs) are increasingly susceptible to sophisticated adversarial attacks, including adaptive strategies specifically designed to bypass existing defenses. To address this vulnerability, we propose MirrorCheck, a robust and model-agnostic detection framework that operates effectively in both unimodal and multimodal settings. MirrorCheck leverages Text-to-Image (T2I) models to regenerate visual content from captions produced by the target model and assesses semantic consistency by comparing feature-space embeddings between the original and synthesized images. To enhance ro"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.09250","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/2406.09250/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":"2406.09250","created_at":"2026-05-25T02:01:00.958584+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.09250v3","created_at":"2026-05-25T02:01:00.958584+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.09250","created_at":"2026-05-25T02:01:00.958584+00:00"},{"alias_kind":"pith_short_12","alias_value":"6GCICKCRNHRT","created_at":"2026-05-25T02:01:00.958584+00:00"},{"alias_kind":"pith_short_16","alias_value":"6GCICKCRNHRTDBWJ","created_at":"2026-05-25T02:01:00.958584+00:00"},{"alias_kind":"pith_short_8","alias_value":"6GCICKCR","created_at":"2026-05-25T02:01:00.958584+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2502.05206","citing_title":"Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety","ref_index":247,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK","json":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK.json","graph_json":"https://pith.science/api/pith-number/6GCICKCRNHRTDBWJHOR6EKAPTK/graph.json","events_json":"https://pith.science/api/pith-number/6GCICKCRNHRTDBWJHOR6EKAPTK/events.json","paper":"https://pith.science/paper/6GCICKCR"},"agent_actions":{"view_html":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK","download_json":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK.json","view_paper":"https://pith.science/paper/6GCICKCR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.09250&json=true","fetch_graph":"https://pith.science/api/pith-number/6GCICKCRNHRTDBWJHOR6EKAPTK/graph.json","fetch_events":"https://pith.science/api/pith-number/6GCICKCRNHRTDBWJHOR6EKAPTK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK/action/storage_attestation","attest_author":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK/action/author_attestation","sign_citation":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK/action/citation_signature","submit_replication":"https://pith.science/pith/6GCICKCRNHRTDBWJHOR6EKAPTK/action/replication_record"}},"created_at":"2026-05-25T02:01:00.958584+00:00","updated_at":"2026-05-25T02:01:00.958584+00:00"}