{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:2QSTNGKZLGTLY2E7MY5LB42EJY","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":"227359a1f72676a27bbec2eddc2df951cb2a0e98f13ea2785adf05f80f112670","cross_cats_sorted":["cs.AI","cs.CL","cs.CR"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-01T13:12:30Z","title_canon_sha256":"b5b7574977280de2790c30b00d824d571ead2e103749ca296b1afd4fc675c647"},"schema_version":"1.0","source":{"id":"2404.01099","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.01099","created_at":"2026-07-05T08:57:01Z"},{"alias_kind":"arxiv_version","alias_value":"2404.01099v2","created_at":"2026-07-05T08:57:01Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.01099","created_at":"2026-07-05T08:57:01Z"},{"alias_kind":"pith_short_12","alias_value":"2QSTNGKZLGTL","created_at":"2026-07-05T08:57:01Z"},{"alias_kind":"pith_short_16","alias_value":"2QSTNGKZLGTLY2E7","created_at":"2026-07-05T08:57:01Z"},{"alias_kind":"pith_short_8","alias_value":"2QSTNGKZ","created_at":"2026-07-05T08:57:01Z"}],"graph_snapshots":[{"event_id":"sha256:330ee297ecc74c9287880fe5349fd5c73b4f911e7e42b39b07a75b215ae066f3","target":"graph","created_at":"2026-07-05T08:57:01Z","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/2404.01099/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Current Large Language Models (LLMs), even those tuned for safety and alignment, are susceptible to jailbreaking. Some have found that just further fine-tuning an aligned model with benign data (i.e., data without harmful content) surprisingly leads to substantial degradation in safety. We delve into the data-centric aspects of why benign fine-tuning inadvertently contributes to jailbreaking. First, we represent fine-tuning data through two lenses: representation and gradient spaces. Additionally, we propose a bi-directional anchoring method that, during the selection process, prioritizes data","authors_text":"Luxi He, Mengzhou Xia, Peter Henderson","cross_cats":["cs.AI","cs.CL","cs.CR"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-01T13:12:30Z","title":"What is in Your Safe Data? Identifying Benign Data that Breaks Safety"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.01099","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:db97cdeb257b2677b6472ea5af07ce04168594e259b0aa94c5d95191d3e5a63f","target":"record","created_at":"2026-07-05T08:57:01Z","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":"227359a1f72676a27bbec2eddc2df951cb2a0e98f13ea2785adf05f80f112670","cross_cats_sorted":["cs.AI","cs.CL","cs.CR"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-04-01T13:12:30Z","title_canon_sha256":"b5b7574977280de2790c30b00d824d571ead2e103749ca296b1afd4fc675c647"},"schema_version":"1.0","source":{"id":"2404.01099","kind":"arxiv","version":2}},"canonical_sha256":"d42536995959a6bc689f663ab0f3444e14f83af0b3568350fe1217f5614e6bbe","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"d42536995959a6bc689f663ab0f3444e14f83af0b3568350fe1217f5614e6bbe","first_computed_at":"2026-07-05T08:57:01.816730Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:57:01.816730Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Yo/ZD9vhFITimPd8d1u+IIGHP/P76BQ4aGZ5DoLtULUZNZ0QyI4GfwwEwGMNlT8/l+Mpqgz2+c4EnQbSX1GeBQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:57:01.817127Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.01099","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:db97cdeb257b2677b6472ea5af07ce04168594e259b0aa94c5d95191d3e5a63f","sha256:330ee297ecc74c9287880fe5349fd5c73b4f911e7e42b39b07a75b215ae066f3"],"state_sha256":"c4164a854351977c1bea2ade26369b1ac54cfa6698db35edb75daef7ed9933bd"}