{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:6PR5XLRBQX6KRHISFIGTVZVTJV","short_pith_number":"pith:6PR5XLRB","schema_version":"1.0","canonical_sha256":"f3e3dbae2185fca89d122a0d3ae6b34d5ed8f410614ad2836b08e94a430c3b62","source":{"kind":"arxiv","id":"2312.10766","version":4},"attestation_state":"computed","paper":{"title":"JailGuard: A Universal Detection Framework for LLM Prompt-based Attacks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Cen Zhang, Chao Shen, Jie Zhang, Ming Hu, Shiqing Ma, Tianlin Li, Xiaojun Jia, Xiaoyu Zhang, Yang Liu, Yihao Huang","submitted_at":"2023-12-17T17:02:14Z","abstract_excerpt":"The systems and software powered by Large Language Models (LLMs) and Multi-Modal LLMs (MLLMs) have played a critical role in numerous scenarios. However, current LLM systems are vulnerable to prompt-based attacks, with jailbreaking attacks enabling the LLM system to generate harmful content, while hijacking attacks manipulate the LLM system to perform attacker-desired tasks, underscoring the necessity for detection tools. Unfortunately, existing detecting approaches are usually tailored to specific attacks, resulting in poor generalization in detecting various attacks across different modaliti"},"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":"2312.10766","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2023-12-17T17:02:14Z","cross_cats_sorted":[],"title_canon_sha256":"49850a54514a5be33a00a4cd60d4a5539915306454b6d73a8da2f2e93ca0f332","abstract_canon_sha256":"247d4165c687b8803ad0835d57dfb1fa0e9e69ba860adc59e8d079a1e63098b7"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:31:40.764759Z","signature_b64":"BPiDClFYjXWRf7uyDnrTVFyJ6rM/z6lPIYXSBj/M3S7ACdvXTImD7aJkCkg6hzYSZtPGEaaxN3Cy7VDMiyKeDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f3e3dbae2185fca89d122a0d3ae6b34d5ed8f410614ad2836b08e94a430c3b62","last_reissued_at":"2026-07-05T10:31:40.764068Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:31:40.764068Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"JailGuard: A Universal Detection Framework for LLM Prompt-based Attacks","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CR","authors_text":"Cen Zhang, Chao Shen, Jie Zhang, Ming Hu, Shiqing Ma, Tianlin Li, Xiaojun Jia, Xiaoyu Zhang, Yang Liu, Yihao Huang","submitted_at":"2023-12-17T17:02:14Z","abstract_excerpt":"The systems and software powered by Large Language Models (LLMs) and Multi-Modal LLMs (MLLMs) have played a critical role in numerous scenarios. However, current LLM systems are vulnerable to prompt-based attacks, with jailbreaking attacks enabling the LLM system to generate harmful content, while hijacking attacks manipulate the LLM system to perform attacker-desired tasks, underscoring the necessity for detection tools. Unfortunately, existing detecting approaches are usually tailored to specific attacks, resulting in poor generalization in detecting various attacks across different modaliti"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2312.10766","kind":"arxiv","version":4},"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/2312.10766/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":"2312.10766","created_at":"2026-07-05T10:31:40.764157+00:00"},{"alias_kind":"arxiv_version","alias_value":"2312.10766v4","created_at":"2026-07-05T10:31:40.764157+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2312.10766","created_at":"2026-07-05T10:31:40.764157+00:00"},{"alias_kind":"pith_short_12","alias_value":"6PR5XLRBQX6K","created_at":"2026-07-05T10:31:40.764157+00:00"},{"alias_kind":"pith_short_16","alias_value":"6PR5XLRBQX6KRHIS","created_at":"2026-07-05T10:31:40.764157+00:00"},{"alias_kind":"pith_short_8","alias_value":"6PR5XLRB","created_at":"2026-07-05T10:31:40.764157+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":6,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.07706","citing_title":"MLingualFC: Evaluating Jailbreak Vulnerabilities in Multilingual Vision-Language Models","ref_index":76,"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":285,"is_internal_anchor":false},{"citing_arxiv_id":"2503.06223","citing_title":"RedDiffuser: Auditing Multimodal Safety Failures in Vision-Language Models via Reinforced Diffusion","ref_index":7,"is_internal_anchor":false},{"citing_arxiv_id":"2605.15377","citing_title":"Ensemble Monitoring for AI Control: Diverse Signals Outweigh More Compute","ref_index":67,"is_internal_anchor":false},{"citing_arxiv_id":"2507.21540","citing_title":"PRISM: Programmatic Reasoning with Image Sequence Manipulation for LVLM Jailbreaking","ref_index":47,"is_internal_anchor":false},{"citing_arxiv_id":"2407.04295","citing_title":"Jailbreak Attacks and Defenses Against Large Language Models: A Survey","ref_index":112,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV","json":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV.json","graph_json":"https://pith.science/api/pith-number/6PR5XLRBQX6KRHISFIGTVZVTJV/graph.json","events_json":"https://pith.science/api/pith-number/6PR5XLRBQX6KRHISFIGTVZVTJV/events.json","paper":"https://pith.science/paper/6PR5XLRB"},"agent_actions":{"view_html":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV","download_json":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV.json","view_paper":"https://pith.science/paper/6PR5XLRB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2312.10766&json=true","fetch_graph":"https://pith.science/api/pith-number/6PR5XLRBQX6KRHISFIGTVZVTJV/graph.json","fetch_events":"https://pith.science/api/pith-number/6PR5XLRBQX6KRHISFIGTVZVTJV/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV/action/timestamp_anchor","attest_storage":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV/action/storage_attestation","attest_author":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV/action/author_attestation","sign_citation":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV/action/citation_signature","submit_replication":"https://pith.science/pith/6PR5XLRBQX6KRHISFIGTVZVTJV/action/replication_record"}},"created_at":"2026-07-05T10:31:40.764157+00:00","updated_at":"2026-07-05T10:31:40.764157+00:00"}