{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:UDVPJ3SOKTTWDP4PMEKS4L3MPB","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":"d97337f48dec872ea6f53b399645d2cab8030048d806b44bf3ee06b6192e8632","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-13T07:36:45Z","title_canon_sha256":"4ca1b23e57245cfac1f4b6bee86a2d52c41b2e50dc6410fabf132361a6753a1b"},"schema_version":"1.0","source":{"id":"2410.09760","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.09760","created_at":"2026-07-05T10:07:45Z"},{"alias_kind":"arxiv_version","alias_value":"2410.09760v3","created_at":"2026-07-05T10:07:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.09760","created_at":"2026-07-05T10:07:45Z"},{"alias_kind":"pith_short_12","alias_value":"UDVPJ3SOKTTW","created_at":"2026-07-05T10:07:45Z"},{"alias_kind":"pith_short_16","alias_value":"UDVPJ3SOKTTWDP4P","created_at":"2026-07-05T10:07:45Z"},{"alias_kind":"pith_short_8","alias_value":"UDVPJ3SO","created_at":"2026-07-05T10:07:45Z"}],"graph_snapshots":[{"event_id":"sha256:836f42f856799141920fb6b8727d49edd674debe25ae536db5a78b19dfdffad6","target":"graph","created_at":"2026-07-05T10:07:45Z","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/2410.09760/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Harmful fine-tuning attack poses a serious threat to the online fine-tuning service. Vaccine, a recent alignment-stage defense, applies uniform perturbation to all layers of embedding to make the model robust to the simulated embedding drift. However, applying layer-wise uniform perturbation may lead to excess perturbations for some particular safety-irrelevant layers, resulting in defense performance degradation and unnecessary memory consumption. To address this limitation, we propose Targeted Vaccine (T-Vaccine), a memory-efficient safety alignment method that applies perturbation to only s","authors_text":"Guozhi Liu, Li Shen, Qi Mu, Ruichao Mo, Tiansheng Huang, Weiwei Lin","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-13T07:36:45Z","title":"Targeted Vaccine: Safety Alignment for Large Language Models against Harmful Fine-Tuning via Layer-wise Perturbation"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.09760","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:d9e8848148f7d9ea44ad0fe0054a3f8a3134f734c6a00367f442fc56635118c4","target":"record","created_at":"2026-07-05T10:07:45Z","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":"d97337f48dec872ea6f53b399645d2cab8030048d806b44bf3ee06b6192e8632","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-13T07:36:45Z","title_canon_sha256":"4ca1b23e57245cfac1f4b6bee86a2d52c41b2e50dc6410fabf132361a6753a1b"},"schema_version":"1.0","source":{"id":"2410.09760","kind":"arxiv","version":3}},"canonical_sha256":"a0eaf4ee4e54e761bf8f61152e2f6c7851b5de788213425f4c417b337911ec92","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a0eaf4ee4e54e761bf8f61152e2f6c7851b5de788213425f4c417b337911ec92","first_computed_at":"2026-07-05T10:07:45.388018Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:07:45.388018Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"aGV/PXLpNmvfziLLLphQHla+Hh6WPHemGnP8JBeUA6h9Z3jRR3CiycUK9SvcZyclK25QpuLxEm15+7WHIfyjDQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:07:45.388453Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.09760","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d9e8848148f7d9ea44ad0fe0054a3f8a3134f734c6a00367f442fc56635118c4","sha256:836f42f856799141920fb6b8727d49edd674debe25ae536db5a78b19dfdffad6"],"state_sha256":"de8b5fcb38992c086f04a204ba1445b412ef4b00c44d4fcbf04106da99c2077e"}