{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TVC6GVW7DXGN2G2KR47KHKWSVL","short_pith_number":"pith:TVC6GVW7","schema_version":"1.0","canonical_sha256":"9d45e356df1dccdd1b4a8f3ea3aad2aaf6188a62f88cf1c5f79fc356c96e6382","source":{"kind":"arxiv","id":"2405.09113","version":2},"attestation_state":"computed","paper":{"title":"Efficient LLM Jailbreak via Adaptive Dense-to-sparse Constrained Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kai Chen, Kai Hu, Lijun Yu, Matt Fredrikson, Tianjun Yao, Weichen Yu, Wenhe Liu, Xiang Li, Yining Li, Zhiqiang Shen","submitted_at":"2024-05-15T06:11:24Z","abstract_excerpt":"Recent research indicates that large language models (LLMs) are susceptible to jailbreaking attacks that can generate harmful content. This paper introduces a novel token-level attack method, Adaptive Dense-to-Sparse Constrained Optimization (ADC), which has been shown to successfully jailbreak multiple open-source LLMs. Drawing inspiration from the difficulties of discrete token optimization, our method relaxes the discrete jailbreak optimization into a continuous optimization process while gradually increasing the sparsity of the optimizing vectors. This technique effectively bridges the gap"},"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":"2405.09113","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-15T06:11:24Z","cross_cats_sorted":[],"title_canon_sha256":"886c1b9dba175745cbe294d2893cc351ec34fe0c5a06dbdb70e17267f0c6e660","abstract_canon_sha256":"0f39411e9d536897329e0049ee43345a4678310e49bb61818304dd636d0e8d01"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:13:00.428061Z","signature_b64":"XCp4uyRT5ezRJLCEAR5qfIinqotxj4T98zn6xtzSoMi+ALtyP08CI9X5/Qj2nNDNEBiysNi++B7YfSmgBUgSBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d45e356df1dccdd1b4a8f3ea3aad2aaf6188a62f88cf1c5f79fc356c96e6382","last_reissued_at":"2026-07-05T10:13:00.427558Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:13:00.427558Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Efficient LLM Jailbreak via Adaptive Dense-to-sparse Constrained Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Kai Chen, Kai Hu, Lijun Yu, Matt Fredrikson, Tianjun Yao, Weichen Yu, Wenhe Liu, Xiang Li, Yining Li, Zhiqiang Shen","submitted_at":"2024-05-15T06:11:24Z","abstract_excerpt":"Recent research indicates that large language models (LLMs) are susceptible to jailbreaking attacks that can generate harmful content. This paper introduces a novel token-level attack method, Adaptive Dense-to-Sparse Constrained Optimization (ADC), which has been shown to successfully jailbreak multiple open-source LLMs. Drawing inspiration from the difficulties of discrete token optimization, our method relaxes the discrete jailbreak optimization into a continuous optimization process while gradually increasing the sparsity of the optimizing vectors. This technique effectively bridges the gap"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.09113","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/2405.09113/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":"2405.09113","created_at":"2026-07-05T10:13:00.427618+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.09113v2","created_at":"2026-07-05T10:13:00.427618+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.09113","created_at":"2026-07-05T10:13:00.427618+00:00"},{"alias_kind":"pith_short_12","alias_value":"TVC6GVW7DXGN","created_at":"2026-07-05T10:13:00.427618+00:00"},{"alias_kind":"pith_short_16","alias_value":"TVC6GVW7DXGN2G2K","created_at":"2026-07-05T10:13:00.427618+00:00"},{"alias_kind":"pith_short_8","alias_value":"TVC6GVW7","created_at":"2026-07-05T10:13:00.427618+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL","json":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL.json","graph_json":"https://pith.science/api/pith-number/TVC6GVW7DXGN2G2KR47KHKWSVL/graph.json","events_json":"https://pith.science/api/pith-number/TVC6GVW7DXGN2G2KR47KHKWSVL/events.json","paper":"https://pith.science/paper/TVC6GVW7"},"agent_actions":{"view_html":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL","download_json":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL.json","view_paper":"https://pith.science/paper/TVC6GVW7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.09113&json=true","fetch_graph":"https://pith.science/api/pith-number/TVC6GVW7DXGN2G2KR47KHKWSVL/graph.json","fetch_events":"https://pith.science/api/pith-number/TVC6GVW7DXGN2G2KR47KHKWSVL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL/action/storage_attestation","attest_author":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL/action/author_attestation","sign_citation":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL/action/citation_signature","submit_replication":"https://pith.science/pith/TVC6GVW7DXGN2G2KR47KHKWSVL/action/replication_record"}},"created_at":"2026-07-05T10:13:00.427618+00:00","updated_at":"2026-07-05T10:13:00.427618+00:00"}