{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:Y7FBCUDMN44B4YX772HLQFNLQC","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"eefd1597a15aef49b72fac7acf03deb50d73e32cb8118ffae4ff8192e0f23c10","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2023-07-26T23:11:15Z","title_canon_sha256":"8252ebc7a01dc4db277837a34671e7341de2863d5a81e66b4cea88ca97ea8172"},"schema_version":"1.0","source":{"id":"2307.14539","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2307.14539","created_at":"2026-07-05T06:59:34Z"},{"alias_kind":"arxiv_version","alias_value":"2307.14539v2","created_at":"2026-07-05T06:59:34Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2307.14539","created_at":"2026-07-05T06:59:34Z"},{"alias_kind":"pith_short_12","alias_value":"Y7FBCUDMN44B","created_at":"2026-07-05T06:59:34Z"},{"alias_kind":"pith_short_16","alias_value":"Y7FBCUDMN44B4YX7","created_at":"2026-07-05T06:59:34Z"},{"alias_kind":"pith_short_8","alias_value":"Y7FBCUDM","created_at":"2026-07-05T06:59:34Z"}],"graph_snapshots":[{"event_id":"sha256:3fbba2934172abaf6875ca2849d8ef453e112e384d49235b6ee882d8d8a88acd","target":"graph","created_at":"2026-07-05T06:59:34Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2307.14539/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce new jailbreak attacks on vision language models (VLMs), which use aligned LLMs and are resilient to text-only jailbreak attacks. Specifically, we develop cross-modality attacks on alignment where we pair adversarial images going through the vision encoder with textual prompts to break the alignment of the language model. Our attacks employ a novel compositional strategy that combines an image, adversarially targeted towards toxic embeddings, with generic prompts to accomplish the jailbreak. Thus, the LLM draws the context to answer the generic prompt from the adversarial image. Th","authors_text":"Erfan Shayegani, Nael Abu-Ghazaleh, Yue Dong","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2023-07-26T23:11:15Z","title":"Jailbreak in pieces: Compositional Adversarial Attacks on Multi-Modal Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2307.14539","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:987f823d39a3a071e9d9c82bcd9dbefa480152e2eb7a355d2f12c55a4c7ac17c","target":"record","created_at":"2026-07-05T06:59:34Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"eefd1597a15aef49b72fac7acf03deb50d73e32cb8118ffae4ff8192e0f23c10","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2023-07-26T23:11:15Z","title_canon_sha256":"8252ebc7a01dc4db277837a34671e7341de2863d5a81e66b4cea88ca97ea8172"},"schema_version":"1.0","source":{"id":"2307.14539","kind":"arxiv","version":2}},"canonical_sha256":"c7ca11506c6f381e62fffe8eb815ab80948316057d4f214d0ffb9c3a908afa26","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c7ca11506c6f381e62fffe8eb815ab80948316057d4f214d0ffb9c3a908afa26","first_computed_at":"2026-07-05T06:59:34.099246Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:59:34.099246Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"elgIl5Nzw8gMY8DTvKx7GbkjFfbWfYrtHo4TP9IWwCA2FuC7RpQVmdXFLp8PFi5QAYteHzbni3v3kVuOWqvFCA==","signature_status":"signed_v1","signed_at":"2026-07-05T06:59:34.099632Z","signed_message":"canonical_sha256_bytes"},"source_id":"2307.14539","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:987f823d39a3a071e9d9c82bcd9dbefa480152e2eb7a355d2f12c55a4c7ac17c","sha256:3fbba2934172abaf6875ca2849d8ef453e112e384d49235b6ee882d8d8a88acd"],"state_sha256":"aea2b73a9ede22dd364cf2f08d87950b1e50c573dc7c32fbac4ad039881ad485"}