{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:YKPYPOCJ7TF247C32R6VAXPX5E","short_pith_number":"pith:YKPYPOCJ","schema_version":"1.0","canonical_sha256":"c29f87b849fccbae7c5bd47d505df7e90da86d723d7d9f2724c93a2a35207640","source":{"kind":"arxiv","id":"2503.15560","version":1},"attestation_state":"computed","paper":{"title":"Temporal Context Awareness: A Defense Framework Against Multi-turn Manipulation Attacks on Large Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CR","authors_text":"Assaf Namer, Prashant Kulkarni","submitted_at":"2025-03-18T22:30:17Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly vulnerable to sophisticated multi-turn manipulation attacks, where adversaries strategically build context through seemingly benign conversational turns to circumvent safety measures and elicit harmful or unauthorized responses. These attacks exploit the temporal nature of dialogue to evade single-turn detection methods, representing a critical security vulnerability with significant implications for real-world deployments.\n  This paper introduces the Temporal Context Awareness (TCA) framework, a novel defense mechanism designed to address this cha"},"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":"2503.15560","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CR","submitted_at":"2025-03-18T22:30:17Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"45f19127ca8e564aecb3ba143f17c59d0f8e8408eb6a0da3f7b33704c45b9825","abstract_canon_sha256":"a5e6eb90bccc7e2266bf8811ae258998906dc3eecf3258ac9f68b0b98c955a84"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:35:52.683092Z","signature_b64":"wlUflwfV2InPV24PCqJiXjZVNUwIodbmvLr3u1Gnv9gHCMH88lDy8PDyDFJHvihud4xGOF1ZWMMaisYFbbpwBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c29f87b849fccbae7c5bd47d505df7e90da86d723d7d9f2724c93a2a35207640","last_reissued_at":"2026-07-05T10:35:52.682562Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:35:52.682562Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Temporal Context Awareness: A Defense Framework Against Multi-turn Manipulation Attacks on Large Language Models","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CR","authors_text":"Assaf Namer, Prashant Kulkarni","submitted_at":"2025-03-18T22:30:17Z","abstract_excerpt":"Large Language Models (LLMs) are increasingly vulnerable to sophisticated multi-turn manipulation attacks, where adversaries strategically build context through seemingly benign conversational turns to circumvent safety measures and elicit harmful or unauthorized responses. These attacks exploit the temporal nature of dialogue to evade single-turn detection methods, representing a critical security vulnerability with significant implications for real-world deployments.\n  This paper introduces the Temporal Context Awareness (TCA) framework, a novel defense mechanism designed to address this cha"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.15560","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/2503.15560/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":"2503.15560","created_at":"2026-07-05T10:35:52.682633+00:00"},{"alias_kind":"arxiv_version","alias_value":"2503.15560v1","created_at":"2026-07-05T10:35:52.682633+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.15560","created_at":"2026-07-05T10:35:52.682633+00:00"},{"alias_kind":"pith_short_12","alias_value":"YKPYPOCJ7TF2","created_at":"2026-07-05T10:35:52.682633+00:00"},{"alias_kind":"pith_short_16","alias_value":"YKPYPOCJ7TF247C3","created_at":"2026-07-05T10:35:52.682633+00:00"},{"alias_kind":"pith_short_8","alias_value":"YKPYPOCJ","created_at":"2026-07-05T10:35:52.682633+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2509.05471","citing_title":"Behind the Mask: Benchmarking Camouflaged Jailbreaks in Large Language Models","ref_index":11,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E","json":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E.json","graph_json":"https://pith.science/api/pith-number/YKPYPOCJ7TF247C32R6VAXPX5E/graph.json","events_json":"https://pith.science/api/pith-number/YKPYPOCJ7TF247C32R6VAXPX5E/events.json","paper":"https://pith.science/paper/YKPYPOCJ"},"agent_actions":{"view_html":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E","download_json":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E.json","view_paper":"https://pith.science/paper/YKPYPOCJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2503.15560&json=true","fetch_graph":"https://pith.science/api/pith-number/YKPYPOCJ7TF247C32R6VAXPX5E/graph.json","fetch_events":"https://pith.science/api/pith-number/YKPYPOCJ7TF247C32R6VAXPX5E/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E/action/storage_attestation","attest_author":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E/action/author_attestation","sign_citation":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E/action/citation_signature","submit_replication":"https://pith.science/pith/YKPYPOCJ7TF247C32R6VAXPX5E/action/replication_record"}},"created_at":"2026-07-05T10:35:52.682633+00:00","updated_at":"2026-07-05T10:35:52.682633+00:00"}