{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:JR5FO6HRNBT6X6EGSOV5SCQLGJ","short_pith_number":"pith:JR5FO6HR","schema_version":"1.0","canonical_sha256":"4c7a5778f16867ebf88693abd90a0b327ae40f56569f936e7cdfd5799c9daf81","source":{"kind":"arxiv","id":"2503.21598","version":1},"attestation_state":"computed","paper":{"title":"Prompt, Divide, and Conquer: Bypassing Large Language Model Safety Filters via Segmented and Distributed Prompt Processing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CR","authors_text":"Ahmed Hussain, Johan Wahr\\'eus, Panos Papadimitratos","submitted_at":"2025-03-27T15:19:55Z","abstract_excerpt":"Large Language Models (LLMs) have transformed task automation and content generation across various domains while incorporating safety filters to prevent misuse. We introduce a novel jailbreaking framework that employs distributed prompt processing combined with iterative refinements to bypass these safety measures, particularly in generating malicious code. Our architecture consists of four key modules: prompt segmentation, parallel processing, response aggregation, and LLM-based jury evaluation. Tested on 500 malicious prompts across 10 cybersecurity categories, the framework achieves a 73.2"},"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":"2503.21598","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-03-27T15:19:55Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"23f44ddc9f985633526b4c9041db15b025fa4a0cb6f354131f26b614ce8f7684","abstract_canon_sha256":"e59d72e5e6f2a366b83045d3d35f6a6d3b55a96e4d99a7759401e39f97179e18"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:41:23.642988Z","signature_b64":"Or6WWxHeugqye2hp6fv8GFFIIUolldypHxuq3qSXYuX/ney8w8cUI3vi/H2dglf27ABBLT1IdbxQJQFWC9vtDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4c7a5778f16867ebf88693abd90a0b327ae40f56569f936e7cdfd5799c9daf81","last_reissued_at":"2026-07-05T10:41:23.642553Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:41:23.642553Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Prompt, Divide, and Conquer: Bypassing Large Language Model Safety Filters via Segmented and Distributed Prompt Processing","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CR","authors_text":"Ahmed Hussain, Johan Wahr\\'eus, Panos Papadimitratos","submitted_at":"2025-03-27T15:19:55Z","abstract_excerpt":"Large Language Models (LLMs) have transformed task automation and content generation across various domains while incorporating safety filters to prevent misuse. We introduce a novel jailbreaking framework that employs distributed prompt processing combined with iterative refinements to bypass these safety measures, particularly in generating malicious code. Our architecture consists of four key modules: prompt segmentation, parallel processing, response aggregation, and LLM-based jury evaluation. Tested on 500 malicious prompts across 10 cybersecurity categories, the framework achieves a 73.2"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.21598","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/2503.21598/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":"2503.21598","created_at":"2026-07-05T10:41:23.642608+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.21598v1","created_at":"2026-07-05T10:41:23.642608+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.21598","created_at":"2026-07-05T10:41:23.642608+00:00"},{"alias_kind":"pith_short_12","alias_value":"JR5FO6HRNBT6","created_at":"2026-07-05T10:41:23.642608+00:00"},{"alias_kind":"pith_short_16","alias_value":"JR5FO6HRNBT6X6EG","created_at":"2026-07-05T10:41:23.642608+00:00"},{"alias_kind":"pith_short_8","alias_value":"JR5FO6HR","created_at":"2026-07-05T10:41:23.642608+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2512.21110","citing_title":"Beyond Context: Large Language Models' Failure to Grasp Users' Intent","ref_index":90,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05630","citing_title":"One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05630","citing_title":"One Turn Too Late: Response-Aware Defense Against Hidden Malicious Intent in Multi-Turn Dialogue","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06552","citing_title":"To Lie or Not to Lie? Investigating The Biased Spread of Global Lies by LLMs","ref_index":37,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ","json":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ.json","graph_json":"https://pith.science/api/pith-number/JR5FO6HRNBT6X6EGSOV5SCQLGJ/graph.json","events_json":"https://pith.science/api/pith-number/JR5FO6HRNBT6X6EGSOV5SCQLGJ/events.json","paper":"https://pith.science/paper/JR5FO6HR"},"agent_actions":{"view_html":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ","download_json":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ.json","view_paper":"https://pith.science/paper/JR5FO6HR","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.21598&json=true","fetch_graph":"https://pith.science/api/pith-number/JR5FO6HRNBT6X6EGSOV5SCQLGJ/graph.json","fetch_events":"https://pith.science/api/pith-number/JR5FO6HRNBT6X6EGSOV5SCQLGJ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ/action/storage_attestation","attest_author":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ/action/author_attestation","sign_citation":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ/action/citation_signature","submit_replication":"https://pith.science/pith/JR5FO6HRNBT6X6EGSOV5SCQLGJ/action/replication_record"}},"created_at":"2026-07-05T10:41:23.642608+00:00","updated_at":"2026-07-05T10:41:23.642608+00:00"}