{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:C4GJL74WECPP5QT6LQAAZBHPQ3","short_pith_number":"pith:C4GJL74W","schema_version":"1.0","canonical_sha256":"170c95ff96209efec27e5c000c84ef86eff9338aa940d7caea28d2d060ec10fd","source":{"kind":"arxiv","id":"2404.02532","version":1},"attestation_state":"computed","paper":{"title":"Learn to Disguise: Avoid Refusal Responses in LLM's Defense via a Multi-agent Attacker-Disguiser Game","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Dongsheng Li, Feng Liu, Hongyan Wu, Qianqiao Xu, Yiping Song, Zhen Huang, Zhiliang Tian","submitted_at":"2024-04-03T07:43:11Z","abstract_excerpt":"With the enhanced performance of large models on natural language processing tasks, potential moral and ethical issues of large models arise. There exist malicious attackers who induce large models to jailbreak and generate information containing illegal, privacy-invasive information through techniques such as prompt engineering. As a result, large models counter malicious attackers' attacks using techniques such as safety alignment. However, the strong defense mechanism of the large model through rejection replies is easily identified by attackers and used to strengthen attackers' capabilitie"},"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":"2404.02532","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2024-04-03T07:43:11Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"249aca43c2d9ceddd653c62ec93d79b9fe3b066706bf6384b3c35cce53c4ee1a","abstract_canon_sha256":"d062cf16f18240e988a0882716ecb4da13393b91c3f64bd9339bd6d00ee17bf2"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:03:57.881505Z","signature_b64":"X80lJPeHdZHR6JpvXJTWjhNh/slvbYObGqhAyPsYFvO46siVbWwFjjsfxcNsGZMi5txSVWDtmHs4L104TpDOCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"170c95ff96209efec27e5c000c84ef86eff9338aa940d7caea28d2d060ec10fd","last_reissued_at":"2026-07-05T08:03:57.881030Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:03:57.881030Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learn to Disguise: Avoid Refusal Responses in LLM's Defense via a Multi-agent Attacker-Disguiser Game","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.AI","authors_text":"Dongsheng Li, Feng Liu, Hongyan Wu, Qianqiao Xu, Yiping Song, Zhen Huang, Zhiliang Tian","submitted_at":"2024-04-03T07:43:11Z","abstract_excerpt":"With the enhanced performance of large models on natural language processing tasks, potential moral and ethical issues of large models arise. There exist malicious attackers who induce large models to jailbreak and generate information containing illegal, privacy-invasive information through techniques such as prompt engineering. As a result, large models counter malicious attackers' attacks using techniques such as safety alignment. However, the strong defense mechanism of the large model through rejection replies is easily identified by attackers and used to strengthen attackers' capabilitie"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.02532","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/2404.02532/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":"2404.02532","created_at":"2026-07-05T08:03:57.881085+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.02532v1","created_at":"2026-07-05T08:03:57.881085+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.02532","created_at":"2026-07-05T08:03:57.881085+00:00"},{"alias_kind":"pith_short_12","alias_value":"C4GJL74WECPP","created_at":"2026-07-05T08:03:57.881085+00:00"},{"alias_kind":"pith_short_16","alias_value":"C4GJL74WECPP5QT6","created_at":"2026-07-05T08:03:57.881085+00:00"},{"alias_kind":"pith_short_8","alias_value":"C4GJL74W","created_at":"2026-07-05T08:03:57.881085+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/C4GJL74WECPP5QT6LQAAZBHPQ3","json":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3.json","graph_json":"https://pith.science/api/pith-number/C4GJL74WECPP5QT6LQAAZBHPQ3/graph.json","events_json":"https://pith.science/api/pith-number/C4GJL74WECPP5QT6LQAAZBHPQ3/events.json","paper":"https://pith.science/paper/C4GJL74W"},"agent_actions":{"view_html":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3","download_json":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3.json","view_paper":"https://pith.science/paper/C4GJL74W","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.02532&json=true","fetch_graph":"https://pith.science/api/pith-number/C4GJL74WECPP5QT6LQAAZBHPQ3/graph.json","fetch_events":"https://pith.science/api/pith-number/C4GJL74WECPP5QT6LQAAZBHPQ3/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3/action/timestamp_anchor","attest_storage":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3/action/storage_attestation","attest_author":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3/action/author_attestation","sign_citation":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3/action/citation_signature","submit_replication":"https://pith.science/pith/C4GJL74WECPP5QT6LQAAZBHPQ3/action/replication_record"}},"created_at":"2026-07-05T08:03:57.881085+00:00","updated_at":"2026-07-05T08:03:57.881085+00:00"}