{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:N7QUGNQ3M6Z5N6ZTWNV7CWQP4K","short_pith_number":"pith:N7QUGNQ3","canonical_record":{"source":{"id":"2508.16406","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-22T14:13:16Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"bb3fd506f80b18cb2c0f86ca73fbf9801872b2e92286aff0e97c19b9c58a7ddd","abstract_canon_sha256":"32eed8967eb54ffc5316cfdcb6f2711d977218c415a364df33028363edc98a2e"},"schema_version":"1.0"},"canonical_sha256":"6fe143361b67b3d6fb33b36bf15a0fe2b9547a49476c5990bd3fe6c978ba1ccf","source":{"kind":"arxiv","id":"2508.16406","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.16406","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"arxiv_version","alias_value":"2508.16406v3","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.16406","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"pith_short_12","alias_value":"N7QUGNQ3M6Z5","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"pith_short_16","alias_value":"N7QUGNQ3M6Z5N6ZT","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"pith_short_8","alias_value":"N7QUGNQ3","created_at":"2026-08-11T01:20:05Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:N7QUGNQ3M6Z5N6ZTWNV7CWQP4K","target":"record","payload":{"canonical_record":{"source":{"id":"2508.16406","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-22T14:13:16Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"bb3fd506f80b18cb2c0f86ca73fbf9801872b2e92286aff0e97c19b9c58a7ddd","abstract_canon_sha256":"32eed8967eb54ffc5316cfdcb6f2711d977218c415a364df33028363edc98a2e"},"schema_version":"1.0"},"canonical_sha256":"6fe143361b67b3d6fb33b36bf15a0fe2b9547a49476c5990bd3fe6c978ba1ccf","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:20:05.104941Z","signature_b64":"Vr+STXYL94Wh1XBMwVgr+NRQAKMGUZC9YQp0euas0w5mg7P1ScHmWEM5EEqzmWrkV0tP36xJSsbhXr2/Fu+zCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6fe143361b67b3d6fb33b36bf15a0fe2b9547a49476c5990bd3fe6c978ba1ccf","last_reissued_at":"2026-08-11T01:20:05.102317Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:20:05.102317Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2508.16406","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-11T01:20:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"O6vX5xSkiKNLatFI46zR584R5C/TVBvwFfpYVmQfuj4B5PSQrPh/9J7u4HPmF2Qm8H2TNP8jUzy8xNVpXuMtBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T14:49:11.361327Z"},"content_sha256":"82a774cbb4315bf89841fb866c70e890ca7cffe16cadce5ea6ada52bd4059f8f","schema_version":"1.0","event_id":"sha256:82a774cbb4315bf89841fb866c70e890ca7cffe16cadce5ea6ada52bd4059f8f"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:N7QUGNQ3M6Z5N6ZTWNV7CWQP4K","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Retrieval-Augmented Defense: Adaptive and Controllable Jailbreak Prevention for Large Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.CR","authors_text":"Bill Byrne, Guangyu Yang, Jingbiao Mei, Jinghong Chen, Weizhe Lin","submitted_at":"2025-08-22T14:13:16Z","abstract_excerpt":"Large Language Models (LLMs) remain vulnerable to jailbreak attacks, which attempt to elicit harmful responses from LLMs. The evolving nature and diversity of these attacks pose many challenges for defense systems, including (1) adaptation to counter emerging attack strategies without costly retraining, and (2) control of the trade-off between safety and utility. To address these challenges, we propose Retrieval-Augmented Defense (RAD), a novel framework for jailbreak detection that incorporates a database of known attack examples into Retrieval-Augmented Generation, which is used to infer the"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.16406","kind":"arxiv","version":3},"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/2508.16406/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-11T01:20:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"dW69+1V9HHH7mBvM7qSNCpDiMV2aAGyCEpxIx+ru9kGh199smnV1TmWx2kUS37ILZRtAAJqtcOLH2gT/BswCBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-20T14:49:11.361665Z"},"content_sha256":"b9d4e9b411d220dc92a43f3a85e6daa84c4b9585ddc2ac5340b7875bb59c7c41","schema_version":"1.0","event_id":"sha256:b9d4e9b411d220dc92a43f3a85e6daa84c4b9585ddc2ac5340b7875bb59c7c41"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K/bundle.json","state_url":"https://pith.science/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-20T14:49:11Z","links":{"resolver":"https://pith.science/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K","bundle":"https://pith.science/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K/bundle.json","state":"https://pith.science/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K/state.json","well_known_bundle":"https://pith.science/.well-known/pith/N7QUGNQ3M6Z5N6ZTWNV7CWQP4K/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:N7QUGNQ3M6Z5N6ZTWNV7CWQP4K","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"32eed8967eb54ffc5316cfdcb6f2711d977218c415a364df33028363edc98a2e","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-22T14:13:16Z","title_canon_sha256":"bb3fd506f80b18cb2c0f86ca73fbf9801872b2e92286aff0e97c19b9c58a7ddd"},"schema_version":"1.0","source":{"id":"2508.16406","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2508.16406","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"arxiv_version","alias_value":"2508.16406v3","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2508.16406","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"pith_short_12","alias_value":"N7QUGNQ3M6Z5","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"pith_short_16","alias_value":"N7QUGNQ3M6Z5N6ZT","created_at":"2026-08-11T01:20:05Z"},{"alias_kind":"pith_short_8","alias_value":"N7QUGNQ3","created_at":"2026-08-11T01:20:05Z"}],"graph_snapshots":[{"event_id":"sha256:b9d4e9b411d220dc92a43f3a85e6daa84c4b9585ddc2ac5340b7875bb59c7c41","target":"graph","created_at":"2026-08-11T01:20:05Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2508.16406/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) remain vulnerable to jailbreak attacks, which attempt to elicit harmful responses from LLMs. The evolving nature and diversity of these attacks pose many challenges for defense systems, including (1) adaptation to counter emerging attack strategies without costly retraining, and (2) control of the trade-off between safety and utility. To address these challenges, we propose Retrieval-Augmented Defense (RAD), a novel framework for jailbreak detection that incorporates a database of known attack examples into Retrieval-Augmented Generation, which is used to infer the","authors_text":"Bill Byrne, Guangyu Yang, Jingbiao Mei, Jinghong Chen, Weizhe Lin","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-22T14:13:16Z","title":"Retrieval-Augmented Defense: Adaptive and Controllable Jailbreak Prevention for Large Language Models"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2508.16406","kind":"arxiv","version":3},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:82a774cbb4315bf89841fb866c70e890ca7cffe16cadce5ea6ada52bd4059f8f","target":"record","created_at":"2026-08-11T01:20:05Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"32eed8967eb54ffc5316cfdcb6f2711d977218c415a364df33028363edc98a2e","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CR","submitted_at":"2025-08-22T14:13:16Z","title_canon_sha256":"bb3fd506f80b18cb2c0f86ca73fbf9801872b2e92286aff0e97c19b9c58a7ddd"},"schema_version":"1.0","source":{"id":"2508.16406","kind":"arxiv","version":3}},"canonical_sha256":"6fe143361b67b3d6fb33b36bf15a0fe2b9547a49476c5990bd3fe6c978ba1ccf","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"6fe143361b67b3d6fb33b36bf15a0fe2b9547a49476c5990bd3fe6c978ba1ccf","first_computed_at":"2026-08-11T01:20:05.102317Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-11T01:20:05.102317Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Vr+STXYL94Wh1XBMwVgr+NRQAKMGUZC9YQp0euas0w5mg7P1ScHmWEM5EEqzmWrkV0tP36xJSsbhXr2/Fu+zCA==","signature_status":"signed_v1","signed_at":"2026-08-11T01:20:05.104941Z","signed_message":"canonical_sha256_bytes"},"source_id":"2508.16406","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:82a774cbb4315bf89841fb866c70e890ca7cffe16cadce5ea6ada52bd4059f8f","sha256:b9d4e9b411d220dc92a43f3a85e6daa84c4b9585ddc2ac5340b7875bb59c7c41"],"state_sha256":"1a1bf7201c2dc149a92b4fadd71154ccb0743c42f011fad6e7c67de40853a015"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"XkfeM038u/qnnYJbmo16FtOhNkfFjrteVLQI0Oe9h+Oe5933a2vT3E77MzvdqISs6ADow4L/6V1hXhl1pFJrBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-20T14:49:11.365249Z","bundle_sha256":"a02b2e042dcdd03c17cb69b4599bc401d8bc4e657ac43cbfa4d352fe5a88455d"}}