{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:QBW76FOVIHXAJTMLDQFPNJZDGL","short_pith_number":"pith:QBW76FOV","schema_version":"1.0","canonical_sha256":"806dff15d541ee04cd8b1c0af6a72332c78cea808acbb0eb2516a5c01b4ac374","source":{"kind":"arxiv","id":"2409.14729","version":2},"attestation_state":"computed","paper":{"title":"PROMPTFUZZ: Harnessing Fuzzing Techniques for Robust Testing of Prompt Injection in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Hanwen Miao, Jiahao Yu, Junzheng Shi, Yangguang Shao","submitted_at":"2024-09-23T06:08:32Z","abstract_excerpt":"Large Language Models (LLMs) have gained widespread use in various applications due to their powerful capability to generate human-like text. However, prompt injection attacks, which involve overwriting a model's original instructions with malicious prompts to manipulate the generated text, have raised significant concerns about the security and reliability of LLMs. Ensuring that LLMs are robust against such attacks is crucial for their deployment in real-world applications, particularly in critical tasks.\n  In this paper, we propose PROMPTFUZZ, a novel testing framework that leverages fuzzing"},"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":"2409.14729","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2024-09-23T06:08:32Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"e46d8081e8f4b234d94c1ffb9afce387fdaccb75fccec0e1fe51402e8a079843","abstract_canon_sha256":"92cbad0996a52ed77ae8b79638bee28480406f7dc18acb46cea9137b6a19fb7e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:44:16.577964Z","signature_b64":"QQ5XJguyh13zwSkxmXBSLfjpAUposcOPQte4eIoSqIpcanRvKQTeSxmQI3zuo40milRi+gsgzyLFb4An8JSACg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"806dff15d541ee04cd8b1c0af6a72332c78cea808acbb0eb2516a5c01b4ac374","last_reissued_at":"2026-07-05T10:44:16.577457Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:44:16.577457Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"PROMPTFUZZ: Harnessing Fuzzing Techniques for Robust Testing of Prompt Injection in LLMs","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Hanwen Miao, Jiahao Yu, Junzheng Shi, Yangguang Shao","submitted_at":"2024-09-23T06:08:32Z","abstract_excerpt":"Large Language Models (LLMs) have gained widespread use in various applications due to their powerful capability to generate human-like text. However, prompt injection attacks, which involve overwriting a model's original instructions with malicious prompts to manipulate the generated text, have raised significant concerns about the security and reliability of LLMs. Ensuring that LLMs are robust against such attacks is crucial for their deployment in real-world applications, particularly in critical tasks.\n  In this paper, we propose PROMPTFUZZ, a novel testing framework that leverages fuzzing"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2409.14729","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/2409.14729/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":"2409.14729","created_at":"2026-07-05T10:44:16.577518+00:00"},{"alias_kind":"arxiv_version","alias_value":"2409.14729v2","created_at":"2026-07-05T10:44:16.577518+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2409.14729","created_at":"2026-07-05T10:44:16.577518+00:00"},{"alias_kind":"pith_short_12","alias_value":"QBW76FOVIHXA","created_at":"2026-07-05T10:44:16.577518+00:00"},{"alias_kind":"pith_short_16","alias_value":"QBW76FOVIHXAJTML","created_at":"2026-07-05T10:44:16.577518+00:00"},{"alias_kind":"pith_short_8","alias_value":"QBW76FOV","created_at":"2026-07-05T10:44:16.577518+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20470","citing_title":"Analyzing Defensive Misdirection Against Model-Guided Automated Attacks on Agentic AI Systems","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2606.20470","citing_title":"Analyzing Defensive Misdirection Against Model-Guided Automated Attacks on Agentic AI Systems","ref_index":28,"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":136,"is_internal_anchor":false},{"citing_arxiv_id":"2510.10073","citing_title":"SecureWebArena: A Holistic Security Evaluation Benchmark for LVLM-based Web Agents","ref_index":61,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL","json":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL.json","graph_json":"https://pith.science/api/pith-number/QBW76FOVIHXAJTMLDQFPNJZDGL/graph.json","events_json":"https://pith.science/api/pith-number/QBW76FOVIHXAJTMLDQFPNJZDGL/events.json","paper":"https://pith.science/paper/QBW76FOV"},"agent_actions":{"view_html":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL","download_json":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL.json","view_paper":"https://pith.science/paper/QBW76FOV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2409.14729&json=true","fetch_graph":"https://pith.science/api/pith-number/QBW76FOVIHXAJTMLDQFPNJZDGL/graph.json","fetch_events":"https://pith.science/api/pith-number/QBW76FOVIHXAJTMLDQFPNJZDGL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL/action/storage_attestation","attest_author":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL/action/author_attestation","sign_citation":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL/action/citation_signature","submit_replication":"https://pith.science/pith/QBW76FOVIHXAJTMLDQFPNJZDGL/action/replication_record"}},"created_at":"2026-07-05T10:44:16.577518+00:00","updated_at":"2026-07-05T10:44:16.577518+00:00"}