{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:JBS6HFKV6WUZZ3LI75EOOHGCLD","short_pith_number":"pith:JBS6HFKV","schema_version":"1.0","canonical_sha256":"4865e39555f5a99ced68ff48e71cc258fc8d116ab625b28da49b58e7cfd51d92","source":{"kind":"arxiv","id":"2405.19668","version":1},"attestation_state":"computed","paper":{"title":"AutoBreach: Universal and Adaptive Jailbreaking with Efficient Wordplay-Guided Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hang Su, Jiawei Chen, Xiao Yang, Yinpeng Dong, Yu Tian, Zhaoxia Yin, Zhengwei Fang","submitted_at":"2024-05-30T03:38:31Z","abstract_excerpt":"Despite the widespread application of large language models (LLMs) across various tasks, recent studies indicate that they are susceptible to jailbreak attacks, which can render their defense mechanisms ineffective. However, previous jailbreak research has frequently been constrained by limited universality, suboptimal efficiency, and a reliance on manual crafting. In response, we rethink the approach to jailbreaking LLMs and formally define three essential properties from the attacker' s perspective, which contributes to guiding the design of jailbreak methods. We further introduce AutoBreach"},"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":"2405.19668","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2024-05-30T03:38:31Z","cross_cats_sorted":[],"title_canon_sha256":"260bd759ddfce45cc41f278394af895b1815c50f69b2847b51ce5f5fb6b399b7","abstract_canon_sha256":"66f17131dff8438f1760701a936faae7639f9b0195b663066cb5ec8454c9ca9c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:25:10.933516Z","signature_b64":"PAZ7wIpze/zNAvSDfBjkvaG9BKPUfHx/v83rIlna8SK3668wV41kOzB8SVt67A6o7XqId3fptQaPMT0fcvzADw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4865e39555f5a99ced68ff48e71cc258fc8d116ab625b28da49b58e7cfd51d92","last_reissued_at":"2026-07-05T08:25:10.933016Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:25:10.933016Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"AutoBreach: Universal and Adaptive Jailbreaking with Efficient Wordplay-Guided Optimization","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Hang Su, Jiawei Chen, Xiao Yang, Yinpeng Dong, Yu Tian, Zhaoxia Yin, Zhengwei Fang","submitted_at":"2024-05-30T03:38:31Z","abstract_excerpt":"Despite the widespread application of large language models (LLMs) across various tasks, recent studies indicate that they are susceptible to jailbreak attacks, which can render their defense mechanisms ineffective. However, previous jailbreak research has frequently been constrained by limited universality, suboptimal efficiency, and a reliance on manual crafting. In response, we rethink the approach to jailbreaking LLMs and formally define three essential properties from the attacker' s perspective, which contributes to guiding the design of jailbreak methods. We further introduce AutoBreach"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.19668","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/2405.19668/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":"2405.19668","created_at":"2026-07-05T08:25:10.933077+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.19668v1","created_at":"2026-07-05T08:25:10.933077+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.19668","created_at":"2026-07-05T08:25:10.933077+00:00"},{"alias_kind":"pith_short_12","alias_value":"JBS6HFKV6WUZ","created_at":"2026-07-05T08:25:10.933077+00:00"},{"alias_kind":"pith_short_16","alias_value":"JBS6HFKV6WUZZ3LI","created_at":"2026-07-05T08:25:10.933077+00:00"},{"alias_kind":"pith_short_8","alias_value":"JBS6HFKV","created_at":"2026-07-05T08:25:10.933077+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.12382","citing_title":"Exploring the Secondary Risks of Large Language Models","ref_index":10,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD","json":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD.json","graph_json":"https://pith.science/api/pith-number/JBS6HFKV6WUZZ3LI75EOOHGCLD/graph.json","events_json":"https://pith.science/api/pith-number/JBS6HFKV6WUZZ3LI75EOOHGCLD/events.json","paper":"https://pith.science/paper/JBS6HFKV"},"agent_actions":{"view_html":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD","download_json":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD.json","view_paper":"https://pith.science/paper/JBS6HFKV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.19668&json=true","fetch_graph":"https://pith.science/api/pith-number/JBS6HFKV6WUZZ3LI75EOOHGCLD/graph.json","fetch_events":"https://pith.science/api/pith-number/JBS6HFKV6WUZZ3LI75EOOHGCLD/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD/action/storage_attestation","attest_author":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD/action/author_attestation","sign_citation":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD/action/citation_signature","submit_replication":"https://pith.science/pith/JBS6HFKV6WUZZ3LI75EOOHGCLD/action/replication_record"}},"created_at":"2026-07-05T08:25:10.933077+00:00","updated_at":"2026-07-05T08:25:10.933077+00:00"}