{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:LZKB3U66RAC665F7YQQRRU3IKK","short_pith_number":"pith:LZKB3U66","schema_version":"1.0","canonical_sha256":"5e541dd3de8805ef74bfc42118d3685285591c17b586406468d37c1f6b8e78de","source":{"kind":"arxiv","id":"2506.12100","version":2},"attestation_state":"computed","paper":{"title":"LLM Embedding-based Attribution (LEA): Quantifying Source Contributions to Generative Model's Response for Vulnerability Analysis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Michael Zuzak, Reza Fayyazi, Shanchieh Jay Yang","submitted_at":"2025-06-12T21:20:10Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly used for cybersecurity threat analysis, but their deployment in security-sensitive environments raises trust and safety concerns. With over 21,000 vulnerabilities disclosed in 2025, manual analysis is infeasible, making scalable and verifiable AI support critical. When querying LLMs, dealing with emerging vulnerabilities is challenging as they have a training cut-off date. While Retrieval-Augmented Generation (RAG) can inject up-to-date context to alleviate the cut-off date limitation, it remains unclear how much LLMs rely on retrieved evidence ver"},"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":"2506.12100","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-06-12T21:20:10Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"8ff9743d17962b835a9de7fc3c698233bde3c3b28cfa81bec2796db7b0995a41","abstract_canon_sha256":"7abdd3c3182ad8b318ea2228473389f72fb8011d2659f61daf19385af144b8b1"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:04:15.720748Z","signature_b64":"S2p3wegcdGJrb3WfbU4eW7pF4sXGlPeIVyYQtEhSF2rqIvWtbRcqs08THP+KZQOAnsSLOuIazp0mXW/sPZ57CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5e541dd3de8805ef74bfc42118d3685285591c17b586406468d37c1f6b8e78de","last_reissued_at":"2026-07-05T12:04:15.720175Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:04:15.720175Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"LLM Embedding-based Attribution (LEA): Quantifying Source Contributions to Generative Model's Response for Vulnerability Analysis","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Michael Zuzak, Reza Fayyazi, Shanchieh Jay Yang","submitted_at":"2025-06-12T21:20:10Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly used for cybersecurity threat analysis, but their deployment in security-sensitive environments raises trust and safety concerns. With over 21,000 vulnerabilities disclosed in 2025, manual analysis is infeasible, making scalable and verifiable AI support critical. When querying LLMs, dealing with emerging vulnerabilities is challenging as they have a training cut-off date. While Retrieval-Augmented Generation (RAG) can inject up-to-date context to alleviate the cut-off date limitation, it remains unclear how much LLMs rely on retrieved evidence ver"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.12100","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/2506.12100/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":"2506.12100","created_at":"2026-07-05T12:04:15.720235+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.12100v2","created_at":"2026-07-05T12:04:15.720235+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.12100","created_at":"2026-07-05T12:04:15.720235+00:00"},{"alias_kind":"pith_short_12","alias_value":"LZKB3U66RAC6","created_at":"2026-07-05T12:04:15.720235+00:00"},{"alias_kind":"pith_short_16","alias_value":"LZKB3U66RAC665F7","created_at":"2026-07-05T12:04:15.720235+00:00"},{"alias_kind":"pith_short_8","alias_value":"LZKB3U66","created_at":"2026-07-05T12:04:15.720235+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK","json":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK.json","graph_json":"https://pith.science/api/pith-number/LZKB3U66RAC665F7YQQRRU3IKK/graph.json","events_json":"https://pith.science/api/pith-number/LZKB3U66RAC665F7YQQRRU3IKK/events.json","paper":"https://pith.science/paper/LZKB3U66"},"agent_actions":{"view_html":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK","download_json":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK.json","view_paper":"https://pith.science/paper/LZKB3U66","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.12100&json=true","fetch_graph":"https://pith.science/api/pith-number/LZKB3U66RAC665F7YQQRRU3IKK/graph.json","fetch_events":"https://pith.science/api/pith-number/LZKB3U66RAC665F7YQQRRU3IKK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK/action/storage_attestation","attest_author":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK/action/author_attestation","sign_citation":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK/action/citation_signature","submit_replication":"https://pith.science/pith/LZKB3U66RAC665F7YQQRRU3IKK/action/replication_record"}},"created_at":"2026-07-05T12:04:15.720235+00:00","updated_at":"2026-07-05T12:04:15.720235+00:00"}