{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:HKGF2EBJAA3OCX2NB2ITG43T62","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":"304e9d3b5e1fcc4474f08e3032ab4ab96f981316108ddbd732c1f5692fb93824","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-04-21T22:18:13Z","title_canon_sha256":"eb84caf788e8759590179a18061b3d31f8915610100ff4da935a05f06edc9f78"},"schema_version":"1.0","source":{"id":"2404.16873","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.16873","created_at":"2026-07-05T11:14:23Z"},{"alias_kind":"arxiv_version","alias_value":"2404.16873v2","created_at":"2026-07-05T11:14:23Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.16873","created_at":"2026-07-05T11:14:23Z"},{"alias_kind":"pith_short_12","alias_value":"HKGF2EBJAA3O","created_at":"2026-07-05T11:14:23Z"},{"alias_kind":"pith_short_16","alias_value":"HKGF2EBJAA3OCX2N","created_at":"2026-07-05T11:14:23Z"},{"alias_kind":"pith_short_8","alias_value":"HKGF2EBJ","created_at":"2026-07-05T11:14:23Z"}],"graph_snapshots":[{"event_id":"sha256:34ee28eeebc728dc38d02aa37d9e8a29de629f76dae072532e4799575bc7b4de","target":"graph","created_at":"2026-07-05T11:14:23Z","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/2404.16873/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are vulnerable to jailbreaking attacks that lead to generation of inappropriate or harmful content. Manual red-teaming requires a time-consuming search for adversarial prompts, whereas automatic adversarial prompt generation often leads to semantically meaningless attacks that do not scale well. In this paper, we present a novel method that uses another LLM, called AdvPrompter, to generate human-readable adversarial prompts in seconds. AdvPrompter, which is trained using an alternating optimization algorithm, generates suffixes that veil the input instruction witho","authors_text":"Anselm Paulus, Arman Zharmagambetov, Brandon Amos, Chuan Guo, Yuandong Tian","cross_cats":["cs.AI","cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-04-21T22:18:13Z","title":"AdvPrompter: Fast Adaptive Adversarial Prompting for LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.16873","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":"sha256:6e9af614b11cf20732529b14b7f7c70e8c8504b1112ea614fa05d9061298af77","target":"record","created_at":"2026-07-05T11:14:23Z","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":"304e9d3b5e1fcc4474f08e3032ab4ab96f981316108ddbd732c1f5692fb93824","cross_cats_sorted":["cs.AI","cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-04-21T22:18:13Z","title_canon_sha256":"eb84caf788e8759590179a18061b3d31f8915610100ff4da935a05f06edc9f78"},"schema_version":"1.0","source":{"id":"2404.16873","kind":"arxiv","version":2}},"canonical_sha256":"3a8c5d10290036e15f4d0e91337373f6bf5cab5382b2c8c558bf55dc07868ad8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3a8c5d10290036e15f4d0e91337373f6bf5cab5382b2c8c558bf55dc07868ad8","first_computed_at":"2026-07-05T11:14:23.462625Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:14:23.462625Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"sQB5/A9O78wrPOpTY8p2XwAuDLxkDviPtlcOVWJeUm4pCdDwpF2GqWR3D/Da03Ueaz3oud1mK8IXL8psR9wSAw==","signature_status":"signed_v1","signed_at":"2026-07-05T11:14:23.463067Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.16873","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:6e9af614b11cf20732529b14b7f7c70e8c8504b1112ea614fa05d9061298af77","sha256:34ee28eeebc728dc38d02aa37d9e8a29de629f76dae072532e4799575bc7b4de"],"state_sha256":"6508cdfadd1d3558b3b8f395a44a1102d3d6abd378fbbcf1b0a47eda78880ae1"}