{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:SNTBQGYEM6ZRN6AIS2RENTMBEO","short_pith_number":"pith:SNTBQGYE","schema_version":"1.0","canonical_sha256":"9366181b0467b316f80896a246cd812380e0ae57f7bd04f6d5c72cbca7a49837","source":{"kind":"arxiv","id":"2402.14857","version":2},"attestation_state":"computed","paper":{"title":"Is the System Message Really Important to Jailbreaks in Large Language Models?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.CL","authors_text":"Ke Li, Xiaotian Zou, Yongkang Chen","submitted_at":"2024-02-20T17:39:40Z","abstract_excerpt":"The rapid evolution of Large Language Models (LLMs) has rendered them indispensable in modern society. While security measures are typically to align LLMs with human values prior to release, recent studies have unveiled a concerning phenomenon named \"Jailbreak\". This term refers to the unexpected and potentially harmful responses generated by LLMs when prompted with malicious questions. Most existing research focus on generating jailbreak prompts but system message configurations vary significantly in experiments. In this paper, we aim to answer a question: Is the system message really importa"},"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":"2402.14857","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-20T17:39:40Z","cross_cats_sorted":["cs.AI","cs.CR"],"title_canon_sha256":"7696119487547e1f0a1f046dfa7da7a685704e25c4c396ca41b1497bbe7bf067","abstract_canon_sha256":"5477b26076a9c0ec1fd3b67bfae38927f060007bd2cb154764fae460a313e790"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:34:13.831285Z","signature_b64":"K0QVZABi4nIU0pnNtO7YEOypgqVjKQ2tkIQGLcmohweGJHa6nKXgxAsLMGWacvIjrGCRIqmF7dwdFf7x1fbBBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9366181b0467b316f80896a246cd812380e0ae57f7bd04f6d5c72cbca7a49837","last_reissued_at":"2026-07-05T08:34:13.830762Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:34:13.830762Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Is the System Message Really Important to Jailbreaks in Large Language Models?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.CR"],"primary_cat":"cs.CL","authors_text":"Ke Li, Xiaotian Zou, Yongkang Chen","submitted_at":"2024-02-20T17:39:40Z","abstract_excerpt":"The rapid evolution of Large Language Models (LLMs) has rendered them indispensable in modern society. While security measures are typically to align LLMs with human values prior to release, recent studies have unveiled a concerning phenomenon named \"Jailbreak\". This term refers to the unexpected and potentially harmful responses generated by LLMs when prompted with malicious questions. Most existing research focus on generating jailbreak prompts but system message configurations vary significantly in experiments. In this paper, we aim to answer a question: Is the system message really importa"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.14857","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/2402.14857/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":"2402.14857","created_at":"2026-07-05T08:34:13.830821+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.14857v2","created_at":"2026-07-05T08:34:13.830821+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.14857","created_at":"2026-07-05T08:34:13.830821+00:00"},{"alias_kind":"pith_short_12","alias_value":"SNTBQGYEM6ZR","created_at":"2026-07-05T08:34:13.830821+00:00"},{"alias_kind":"pith_short_16","alias_value":"SNTBQGYEM6ZRN6AI","created_at":"2026-07-05T08:34:13.830821+00:00"},{"alias_kind":"pith_short_8","alias_value":"SNTBQGYE","created_at":"2026-07-05T08:34:13.830821+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.09125","citing_title":"Unveiling Privacy Risks in Multi-modal Large Language Models: Task-specific Vulnerabilities and Mitigation Challenges","ref_index":113,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30783","citing_title":"Security--Fidelity Tradeoffs: The Hidden Cost of Prompt Injection Defense","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2502.05206","citing_title":"Safety at Scale: A Comprehensive Survey of Large Model and Agent Safety","ref_index":76,"is_internal_anchor":false},{"citing_arxiv_id":"2508.04204","citing_title":"ReasoningGuard: Safeguarding Large Reasoning Models with Inference-time Safety Aha Moments","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2407.04295","citing_title":"Jailbreak Attacks and Defenses Against Large Language Models: A Survey","ref_index":126,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO","json":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO.json","graph_json":"https://pith.science/api/pith-number/SNTBQGYEM6ZRN6AIS2RENTMBEO/graph.json","events_json":"https://pith.science/api/pith-number/SNTBQGYEM6ZRN6AIS2RENTMBEO/events.json","paper":"https://pith.science/paper/SNTBQGYE"},"agent_actions":{"view_html":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO","download_json":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO.json","view_paper":"https://pith.science/paper/SNTBQGYE","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.14857&json=true","fetch_graph":"https://pith.science/api/pith-number/SNTBQGYEM6ZRN6AIS2RENTMBEO/graph.json","fetch_events":"https://pith.science/api/pith-number/SNTBQGYEM6ZRN6AIS2RENTMBEO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO/action/storage_attestation","attest_author":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO/action/author_attestation","sign_citation":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO/action/citation_signature","submit_replication":"https://pith.science/pith/SNTBQGYEM6ZRN6AIS2RENTMBEO/action/replication_record"}},"created_at":"2026-07-05T08:34:13.830821+00:00","updated_at":"2026-07-05T08:34:13.830821+00:00"}