{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:6YNTGGAJUKUDET3VYCCGAQIBHL","short_pith_number":"pith:6YNTGGAJ","canonical_record":{"source":{"id":"2502.11647","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-02-17T10:39:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7a19e990d98fc1eecb61298cc27b2ff7cf6fc286e4d9dcbbd2736fafa0d7b1f9","abstract_canon_sha256":"fb3dd555bb0f7acfecb580687247ddf03717da24ad99255a13d32108edff9c48"},"schema_version":"1.0"},"canonical_sha256":"f61b331809a2a8324f75c0846041013ae74da85d0546390a6e2dd5bdc1bf8e5c","source":{"kind":"arxiv","id":"2502.11647","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.11647","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"arxiv_version","alias_value":"2502.11647v2","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.11647","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"pith_short_12","alias_value":"6YNTGGAJUKUD","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"pith_short_16","alias_value":"6YNTGGAJUKUDET3V","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"pith_short_8","alias_value":"6YNTGGAJ","created_at":"2026-07-05T11:11:41Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:6YNTGGAJUKUDET3VYCCGAQIBHL","target":"record","payload":{"canonical_record":{"source":{"id":"2502.11647","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-02-17T10:39:21Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"7a19e990d98fc1eecb61298cc27b2ff7cf6fc286e4d9dcbbd2736fafa0d7b1f9","abstract_canon_sha256":"fb3dd555bb0f7acfecb580687247ddf03717da24ad99255a13d32108edff9c48"},"schema_version":"1.0"},"canonical_sha256":"f61b331809a2a8324f75c0846041013ae74da85d0546390a6e2dd5bdc1bf8e5c","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:11:41.538005Z","signature_b64":"02GB5TMXcmw+xcS4Rb7bGLoI508YGFAM3jO+nVUkjg8Gpd58JNuK2Q3QOubeVzYVyf3QWtdK81Xn4f8lQHwWDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f61b331809a2a8324f75c0846041013ae74da85d0546390a6e2dd5bdc1bf8e5c","last_reissued_at":"2026-07-05T11:11:41.537492Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:11:41.537492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.11647","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:11:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5uFFHkzqQ8jm9yKldEtl7FP0WmLr6EZg2f/nhHS9HoPI6ju1jZeDsGntcJOrHGylYM5A8ULEV8OgEhkoX9u+Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-12T19:52:58.893461Z"},"content_sha256":"effa862bbde9045de014a30d2ca53a648b9cd799be29ddfd58b7d6f41e423fa1","schema_version":"1.0","event_id":"sha256:effa862bbde9045de014a30d2ca53a648b9cd799be29ddfd58b7d6f41e423fa1"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:6YNTGGAJUKUDET3VYCCGAQIBHL","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DELMAN: Dynamic Defense Against Large Language Model Jailbreaking with Model Editing","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CR","authors_text":"Fenghua Weng, Minlie Huang, Sibei Yang, Wenjie Wang, Yi Wang, Zhan Qin","submitted_at":"2025-02-17T10:39:21Z","abstract_excerpt":"Large Language Models (LLMs) are widely applied in decision making, but their deployment is threatened by jailbreak attacks, where adversarial users manipulate model behavior to bypass safety measures. Existing defense mechanisms, such as safety fine-tuning and model editing, either require extensive parameter modifications or lack precision, leading to performance degradation on general tasks, which is unsuitable to post-deployment safety alignment. To address these challenges, we propose DELMAN (Dynamic Editing for LLMs JAilbreak DefeNse), a novel approach leveraging direct model editing for"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.11647","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/2502.11647/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":"2026-07-05T11:11:41Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Im1G3UdZa9Sptq/KueEHp1x3pu1OSx1d6eD6PSimzEdUn2PMovepWZ3jweVEwpRTRr5DgAabXijzgMTY17CCCA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-12T19:52:58.894126Z"},"content_sha256":"aea4f315dc532b47e337a635269accb2a73f5e5401fb08d5d3514de5e2926ab6","schema_version":"1.0","event_id":"sha256:aea4f315dc532b47e337a635269accb2a73f5e5401fb08d5d3514de5e2926ab6"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6YNTGGAJUKUDET3VYCCGAQIBHL/bundle.json","state_url":"https://pith.science/pith/6YNTGGAJUKUDET3VYCCGAQIBHL/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6YNTGGAJUKUDET3VYCCGAQIBHL/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":"2026-07-12T19:52:58Z","links":{"resolver":"https://pith.science/pith/6YNTGGAJUKUDET3VYCCGAQIBHL","bundle":"https://pith.science/pith/6YNTGGAJUKUDET3VYCCGAQIBHL/bundle.json","state":"https://pith.science/pith/6YNTGGAJUKUDET3VYCCGAQIBHL/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6YNTGGAJUKUDET3VYCCGAQIBHL/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:6YNTGGAJUKUDET3VYCCGAQIBHL","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":"fb3dd555bb0f7acfecb580687247ddf03717da24ad99255a13d32108edff9c48","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-02-17T10:39:21Z","title_canon_sha256":"7a19e990d98fc1eecb61298cc27b2ff7cf6fc286e4d9dcbbd2736fafa0d7b1f9"},"schema_version":"1.0","source":{"id":"2502.11647","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.11647","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"arxiv_version","alias_value":"2502.11647v2","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.11647","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"pith_short_12","alias_value":"6YNTGGAJUKUD","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"pith_short_16","alias_value":"6YNTGGAJUKUDET3V","created_at":"2026-07-05T11:11:41Z"},{"alias_kind":"pith_short_8","alias_value":"6YNTGGAJ","created_at":"2026-07-05T11:11:41Z"}],"graph_snapshots":[{"event_id":"sha256:aea4f315dc532b47e337a635269accb2a73f5e5401fb08d5d3514de5e2926ab6","target":"graph","created_at":"2026-07-05T11:11:41Z","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/2502.11647/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models (LLMs) are widely applied in decision making, but their deployment is threatened by jailbreak attacks, where adversarial users manipulate model behavior to bypass safety measures. Existing defense mechanisms, such as safety fine-tuning and model editing, either require extensive parameter modifications or lack precision, leading to performance degradation on general tasks, which is unsuitable to post-deployment safety alignment. To address these challenges, we propose DELMAN (Dynamic Editing for LLMs JAilbreak DefeNse), a novel approach leveraging direct model editing for","authors_text":"Fenghua Weng, Minlie Huang, Sibei Yang, Wenjie Wang, Yi Wang, Zhan Qin","cross_cats":["cs.AI"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-02-17T10:39:21Z","title":"DELMAN: Dynamic Defense Against Large Language Model Jailbreaking with Model Editing"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.11647","kind":"arxiv","version":2},"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:effa862bbde9045de014a30d2ca53a648b9cd799be29ddfd58b7d6f41e423fa1","target":"record","created_at":"2026-07-05T11:11:41Z","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":"fb3dd555bb0f7acfecb580687247ddf03717da24ad99255a13d32108edff9c48","cross_cats_sorted":["cs.AI"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CR","submitted_at":"2025-02-17T10:39:21Z","title_canon_sha256":"7a19e990d98fc1eecb61298cc27b2ff7cf6fc286e4d9dcbbd2736fafa0d7b1f9"},"schema_version":"1.0","source":{"id":"2502.11647","kind":"arxiv","version":2}},"canonical_sha256":"f61b331809a2a8324f75c0846041013ae74da85d0546390a6e2dd5bdc1bf8e5c","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f61b331809a2a8324f75c0846041013ae74da85d0546390a6e2dd5bdc1bf8e5c","first_computed_at":"2026-07-05T11:11:41.537492Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:11:41.537492Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"02GB5TMXcmw+xcS4Rb7bGLoI508YGFAM3jO+nVUkjg8Gpd58JNuK2Q3QOubeVzYVyf3QWtdK81Xn4f8lQHwWDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:11:41.538005Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.11647","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:effa862bbde9045de014a30d2ca53a648b9cd799be29ddfd58b7d6f41e423fa1","sha256:aea4f315dc532b47e337a635269accb2a73f5e5401fb08d5d3514de5e2926ab6"],"state_sha256":"5c8f685a9a5323288aecf8a2b90410f8e0c3aa6ae7acc11067f46e9efb113f18"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LAi/2DAVQFmpe/Vh+SdSmBQSgHNrn3tBUIYT1zwi0dtkUALLF2YI4PQp9PWm6nvMONyIAY0ieHdz/6Br//U0DA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-12T19:52:58.898093Z","bundle_sha256":"b415e8616904a49faca8bceaa365d79d87e9473ebe0e00355f66f0aaaeb8cf83"}}