{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TEE7T34Z2WIZ5ZFHJOLDT5WXP2","short_pith_number":"pith:TEE7T34Z","schema_version":"1.0","canonical_sha256":"9909f9ef99d5919ee4a74b9639f6d77ea76c54b7e7988fbc6e08298e5538a4b2","source":{"kind":"arxiv","id":"2411.09259","version":2},"attestation_state":"computed","paper":{"title":"Jailbreak Attacks and Defenses against Multimodal Generative Models: A Survey","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Huaibo Huang, Miaoxuan Zhang, Peipei Li, Ran He, Shuhan Xia, Xing Cui, Xuannan Liu, Yueying Zou, Zekun Li","submitted_at":"2024-11-14T07:51:51Z","abstract_excerpt":"The rapid evolution of multimodal foundation models has led to significant advancements in cross-modal understanding and generation across diverse modalities, including text, images, audio, and video. However, these models remain susceptible to jailbreak attacks, which can bypass built-in safety mechanisms and induce the production of potentially harmful content. Consequently, understanding the methods of jailbreak attacks and existing defense mechanisms is essential to ensure the safe deployment of multimodal generative models in real-world scenarios, particularly in security-sensitive applic"},"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":"2411.09259","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-11-14T07:51:51Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"fb377d9f9ee8f79640653dc451c3ef76ca35c580e24aecaae52289fe488ea289","abstract_canon_sha256":"f00fc61302e0f74a58eda7f3d1f24074a88b3fb19c4ba1b23c237a4d82b3e4ee"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:46:31.163232Z","signature_b64":"jPK84isgpZ1C2Si+0KMrTialMeVdHYK5v6app7YMqBCl7jcML6NHL73BdW686loppHiFRQBhirC0LGW4nfHwCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9909f9ef99d5919ee4a74b9639f6d77ea76c54b7e7988fbc6e08298e5538a4b2","last_reissued_at":"2026-07-05T09:46:31.162732Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:46:31.162732Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Jailbreak Attacks and Defenses against Multimodal Generative Models: A Survey","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CV","authors_text":"Huaibo Huang, Miaoxuan Zhang, Peipei Li, Ran He, Shuhan Xia, Xing Cui, Xuannan Liu, Yueying Zou, Zekun Li","submitted_at":"2024-11-14T07:51:51Z","abstract_excerpt":"The rapid evolution of multimodal foundation models has led to significant advancements in cross-modal understanding and generation across diverse modalities, including text, images, audio, and video. However, these models remain susceptible to jailbreak attacks, which can bypass built-in safety mechanisms and induce the production of potentially harmful content. Consequently, understanding the methods of jailbreak attacks and existing defense mechanisms is essential to ensure the safe deployment of multimodal generative models in real-world scenarios, particularly in security-sensitive applic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.09259","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/2411.09259/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":"2411.09259","created_at":"2026-07-05T09:46:31.162788+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.09259v2","created_at":"2026-07-05T09:46:31.162788+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.09259","created_at":"2026-07-05T09:46:31.162788+00:00"},{"alias_kind":"pith_short_12","alias_value":"TEE7T34Z2WIZ","created_at":"2026-07-05T09:46:31.162788+00:00"},{"alias_kind":"pith_short_16","alias_value":"TEE7T34Z2WIZ5ZFH","created_at":"2026-07-05T09:46:31.162788+00:00"},{"alias_kind":"pith_short_8","alias_value":"TEE7T34Z","created_at":"2026-07-05T09:46:31.162788+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.24081","citing_title":"PixJail: Self-Evolving Paper-to-Pipeline Reproduction for Text-to-Image Jailbreak Evaluation","ref_index":11,"is_internal_anchor":false},{"citing_arxiv_id":"2605.18104","citing_title":"Safety Geometry Collapse in Multimodal LLMs and Adaptive Drift Correction","ref_index":53,"is_internal_anchor":false},{"citing_arxiv_id":"2605.00874","citing_title":"Latent Space Probing for Adult Content Detection in Video Generative Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.09253","citing_title":"Mosaic: Multimodal Jailbreak against Closed-Source VLMs via Multi-View Ensemble Optimization","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2","json":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2.json","graph_json":"https://pith.science/api/pith-number/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/graph.json","events_json":"https://pith.science/api/pith-number/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/events.json","paper":"https://pith.science/paper/TEE7T34Z"},"agent_actions":{"view_html":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2","download_json":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2.json","view_paper":"https://pith.science/paper/TEE7T34Z","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.09259&json=true","fetch_graph":"https://pith.science/api/pith-number/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/graph.json","fetch_events":"https://pith.science/api/pith-number/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/action/storage_attestation","attest_author":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/action/author_attestation","sign_citation":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/action/citation_signature","submit_replication":"https://pith.science/pith/TEE7T34Z2WIZ5ZFHJOLDT5WXP2/action/replication_record"}},"created_at":"2026-07-05T09:46:31.162788+00:00","updated_at":"2026-07-05T09:46:31.162788+00:00"}