{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:OMDURBYW7H33GQZRTSXKSVGAQ2","short_pith_number":"pith:OMDURBYW","schema_version":"1.0","canonical_sha256":"7307488716f9f7b343319caea954c08685ae207659bfd3dfe2e464c13d871754","source":{"kind":"arxiv","id":"2406.11036","version":1},"attestation_state":"computed","paper":{"title":"garak: A Framework for Security Probing Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.CL","authors_text":"Erick Galinkin, Jeffrey Martin, Leon Derczynski, Nanna Inie, Subho Majumdar","submitted_at":"2024-06-16T18:18:43Z","abstract_excerpt":"As Large Language Models (LLMs) are deployed and integrated into thousands of applications, the need for scalable evaluation of how models respond to adversarial attacks grows rapidly. However, LLM security is a moving target: models produce unpredictable output, are constantly updated, and the potential adversary is highly diverse: anyone with access to the internet and a decent command of natural language. Further, what constitutes a security weak in one context may not be an issue in a different context; one-fits-all guardrails remain theoretical. In this paper, we argue that it is time to "},"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":"2406.11036","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-06-16T18:18:43Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"d3259b57b76c8e43d41d0c1fc7d672a61c7137171831a17454d4f2a1bff5d38a","abstract_canon_sha256":"9219adae2bd8713985125eccca945c7bb2f6fbc2cf29477e557fbaa4a63f0ac3"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:32:47.954217Z","signature_b64":"8pt9zn/YpnBom8YN/6XsNla5mI54dnodrRH3zo8bv+eF9ujWLeiBV3xm9A5AfI6nH2cAfMqMULgT9gSohllDAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7307488716f9f7b343319caea954c08685ae207659bfd3dfe2e464c13d871754","last_reissued_at":"2026-07-05T08:32:47.953740Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:32:47.953740Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"garak: A Framework for Security Probing Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.CL","authors_text":"Erick Galinkin, Jeffrey Martin, Leon Derczynski, Nanna Inie, Subho Majumdar","submitted_at":"2024-06-16T18:18:43Z","abstract_excerpt":"As Large Language Models (LLMs) are deployed and integrated into thousands of applications, the need for scalable evaluation of how models respond to adversarial attacks grows rapidly. However, LLM security is a moving target: models produce unpredictable output, are constantly updated, and the potential adversary is highly diverse: anyone with access to the internet and a decent command of natural language. Further, what constitutes a security weak in one context may not be an issue in a different context; one-fits-all guardrails remain theoretical. In this paper, we argue that it is time to "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11036","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/2406.11036/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":"2406.11036","created_at":"2026-07-05T08:32:47.953798+00:00"},{"alias_kind":"arxiv_version","alias_value":"2406.11036v1","created_at":"2026-07-05T08:32:47.953798+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11036","created_at":"2026-07-05T08:32:47.953798+00:00"},{"alias_kind":"pith_short_12","alias_value":"OMDURBYW7H33","created_at":"2026-07-05T08:32:47.953798+00:00"},{"alias_kind":"pith_short_16","alias_value":"OMDURBYW7H33GQZR","created_at":"2026-07-05T08:32:47.953798+00:00"},{"alias_kind":"pith_short_8","alias_value":"OMDURBYW","created_at":"2026-07-05T08:32:47.953798+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":17,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26057","citing_title":"The Unfireable Safety Kernel: Execution-Time AI Alignment for AI Agents and Other Escapable AI Systems","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.25476","citing_title":"A Red Teaming Framework for Large Language Models: A Case Study on Faithfulness Evaluation","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.23927","citing_title":"RIFT-Bench: Dynamic Red-teaming For Agentic AI Systems","ref_index":2,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07833","citing_title":"Beyond Pass/Fail: Using Process Mining to Understand How LLMs Resist (and Fail) Red Team Attacks","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2606.03136","citing_title":"PsychoPass: Geometric Profiling of Multi-Turn Adversarial LLM Conversations","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2606.02822","citing_title":"Which Defense Closes Which Threat? Attributing OWASP-LLM-Top-10 Coverage and Its Brittleness Under Paraphrasing","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31639","citing_title":"A Lifecycle and Application-Stack Survey of Large Language Model Vulnerabilities: Attacks, Risks, Defenses, and Open Problems","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31227","citing_title":"Securing the AI Agent: A Unified Framework for Multi-Layer Agent Red Teaming","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2606.31755","citing_title":"A Technical Typology of AI Systems in Public Administration","ref_index":260,"is_internal_anchor":false},{"citing_arxiv_id":"2605.25376","citing_title":"KYA: A Framework-Agnostic Trust Layer for Autonomous Systems with Verifiable Provenance and Hierarchical Policy Composition","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2605.29659","citing_title":"Opir: Efficient Multi-Task Safety Classification for Toxicity, Jailbreaks, Hate Speech, and Harmful Content","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06669","citing_title":"Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2512.05929","citing_title":"LLM Harms: A Taxonomy and Discussion","ref_index":246,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06669","citing_title":"Evaluating Prompt Injection Defenses for Educational LLM Tutors: Security-Usability-Latency Trade-offs","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2605.04019","citing_title":"Redefining AI Red Teaming in the Agentic Era: From Weeks to Hours","ref_index":4,"is_internal_anchor":false},{"citing_arxiv_id":"2605.05058","citing_title":"SoK: Robustness in Large Language Models against Jailbreak Attacks","ref_index":17,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20833","citing_title":"AVISE: Framework for Evaluating the Security of AI Systems","ref_index":2,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2","json":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2.json","graph_json":"https://pith.science/api/pith-number/OMDURBYW7H33GQZRTSXKSVGAQ2/graph.json","events_json":"https://pith.science/api/pith-number/OMDURBYW7H33GQZRTSXKSVGAQ2/events.json","paper":"https://pith.science/paper/OMDURBYW"},"agent_actions":{"view_html":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2","download_json":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2.json","view_paper":"https://pith.science/paper/OMDURBYW","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2406.11036&json=true","fetch_graph":"https://pith.science/api/pith-number/OMDURBYW7H33GQZRTSXKSVGAQ2/graph.json","fetch_events":"https://pith.science/api/pith-number/OMDURBYW7H33GQZRTSXKSVGAQ2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2/action/storage_attestation","attest_author":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2/action/author_attestation","sign_citation":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2/action/citation_signature","submit_replication":"https://pith.science/pith/OMDURBYW7H33GQZRTSXKSVGAQ2/action/replication_record"}},"created_at":"2026-07-05T08:32:47.953798+00:00","updated_at":"2026-07-05T08:32:47.953798+00:00"}