{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:U3HKKR6CCXWTFQFHBVPR7DV6YK","short_pith_number":"pith:U3HKKR6C","schema_version":"1.0","canonical_sha256":"a6cea547c215ed32c0a70d5f1f8ebec297b0badbb5a292824f7b33f3e096d8b4","source":{"kind":"arxiv","id":"2406.03805","version":1},"attestation_state":"computed","paper":{"title":"AutoJailbreak: Exploring Jailbreak Attacks and Defenses through a Dependency Lens","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Hai Yan, Jiawen Shi, Lin Lu, Pan Zhou, Pin-Yu Chen, Wenqi Wei, Zenghui Yuan","submitted_at":"2024-06-06T07:24:41Z","abstract_excerpt":"Jailbreak attacks in large language models (LLMs) entail inducing the models to generate content that breaches ethical and legal norm through the use of malicious prompts, posing a substantial threat to LLM security. Current strategies for jailbreak attack and defense often focus on optimizing locally within specific algorithmic frameworks, resulting in ineffective optimization and limited scalability. In this paper, we present a systematic analysis of the dependency relationships in jailbreak attack and defense techniques, generalizing them to all possible attack surfaces. We employ directed "},"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":"2406.03805","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2024-06-06T07:24:41Z","cross_cats_sorted":[],"title_canon_sha256":"d007146758cc948aecbe4679bba995c00b70e11739b7dc42e8f7c3cef883d6eb","abstract_canon_sha256":"d00cd907b3ec98d092415c46c82386957f2a92cad717c7bffbb9013bce07cf16"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:28:16.186774Z","signature_b64":"ZRgYDswd8rRaSYGyQChRuV1+RxpFJGZqvqUVo2gMqgPggRAVvM7In+azIoNr3OWYDYOuTxqhZ/m9h6508WRwAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a6cea547c215ed32c0a70d5f1f8ebec297b0badbb5a292824f7b33f3e096d8b4","last_reissued_at":"2026-07-05T08:28:16.186295Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:28:16.186295Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AutoJailbreak: Exploring Jailbreak Attacks and Defenses through a Dependency Lens","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Hai Yan, Jiawen Shi, Lin Lu, Pan Zhou, Pin-Yu Chen, Wenqi Wei, Zenghui Yuan","submitted_at":"2024-06-06T07:24:41Z","abstract_excerpt":"Jailbreak attacks in large language models (LLMs) entail inducing the models to generate content that breaches ethical and legal norm through the use of malicious prompts, posing a substantial threat to LLM security. Current strategies for jailbreak attack and defense often focus on optimizing locally within specific algorithmic frameworks, resulting in ineffective optimization and limited scalability. In this paper, we present a systematic analysis of the dependency relationships in jailbreak attack and defense techniques, generalizing them to all possible attack surfaces. We employ directed "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.03805","kind":"arxiv","version":1},"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/2406.03805/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":"2406.03805","created_at":"2026-07-05T08:28:16.186363+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.03805v1","created_at":"2026-07-05T08:28:16.186363+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.03805","created_at":"2026-07-05T08:28:16.186363+00:00"},{"alias_kind":"pith_short_12","alias_value":"U3HKKR6CCXWT","created_at":"2026-07-05T08:28:16.186363+00:00"},{"alias_kind":"pith_short_16","alias_value":"U3HKKR6CCXWTFQFH","created_at":"2026-07-05T08:28:16.186363+00:00"},{"alias_kind":"pith_short_8","alias_value":"U3HKKR6C","created_at":"2026-07-05T08:28:16.186363+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2502.05206","citing_title":"Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety","ref_index":117,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12232","citing_title":"TEMPLATEFUZZ: Fine-Grained Chat Template Fuzzing for Jailbreaking and Red Teaming LLMs","ref_index":35,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK","json":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK.json","graph_json":"https://pith.science/api/pith-number/U3HKKR6CCXWTFQFHBVPR7DV6YK/graph.json","events_json":"https://pith.science/api/pith-number/U3HKKR6CCXWTFQFHBVPR7DV6YK/events.json","paper":"https://pith.science/paper/U3HKKR6C"},"agent_actions":{"view_html":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK","download_json":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK.json","view_paper":"https://pith.science/paper/U3HKKR6C","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.03805&json=true","fetch_graph":"https://pith.science/api/pith-number/U3HKKR6CCXWTFQFHBVPR7DV6YK/graph.json","fetch_events":"https://pith.science/api/pith-number/U3HKKR6CCXWTFQFHBVPR7DV6YK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK/action/storage_attestation","attest_author":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK/action/author_attestation","sign_citation":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK/action/citation_signature","submit_replication":"https://pith.science/pith/U3HKKR6CCXWTFQFHBVPR7DV6YK/action/replication_record"}},"created_at":"2026-07-05T08:28:16.186363+00:00","updated_at":"2026-07-05T08:28:16.186363+00:00"}