{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:YKH7BFTMRYGQRT2GW36XLQKOS5","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":"4a581e9522c89912cb6b850e11ecb4ac20b21c7e097f2126d1e909d294ec9a86","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T06:39:22Z","title_canon_sha256":"13ee7bd8a53f281f32b4eed8928919a7829518e08146142629a4abc9626e7c3e"},"schema_version":"1.0","source":{"id":"2502.09674","kind":"arxiv","version":4}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.09674","created_at":"2026-07-05T11:10:04Z"},{"alias_kind":"arxiv_version","alias_value":"2502.09674v4","created_at":"2026-07-05T11:10:04Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.09674","created_at":"2026-07-05T11:10:04Z"},{"alias_kind":"pith_short_12","alias_value":"YKH7BFTMRYGQ","created_at":"2026-07-05T11:10:04Z"},{"alias_kind":"pith_short_16","alias_value":"YKH7BFTMRYGQRT2G","created_at":"2026-07-05T11:10:04Z"},{"alias_kind":"pith_short_8","alias_value":"YKH7BFTM","created_at":"2026-07-05T11:10:04Z"}],"graph_snapshots":[{"event_id":"sha256:b323ec3099154d35923abc0edd9f88a96a0afbbab74d41bab5db5db75491e955","target":"graph","created_at":"2026-07-05T11:10:04Z","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/2502.09674/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large Language Models' safety-aligned behaviors, such as refusing harmful queries, can be represented by linear directions in activation space. Previous research modeled safety behavior with a single direction, limiting mechanistic understanding to an isolated safety feature. In this work, we discover that safety-aligned behavior is jointly controlled by multi-dimensional directions. Namely, we study the vector space of representation shifts during safety fine-tuning on Llama 3 8B for refusing jailbreaks. By studying orthogonal directions in the space, we first find that a dominant direction g","authors_text":"Haining Yu, Qiguang Chen, Wenbo Pan, Xiangyang Zhou, Xiaohua Jia, Zhichao Liu","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T06:39:22Z","title":"The Hidden Dimensions of LLM Alignment: A Multi-Dimensional Analysis of Orthogonal Safety Directions"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.09674","kind":"arxiv","version":4},"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:d9a01e0c74e3ab5d5427bb331cfc1aba5cd54da32d8860868acf4ac6ea87ba66","target":"record","created_at":"2026-07-05T11:10:04Z","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":"4a581e9522c89912cb6b850e11ecb4ac20b21c7e097f2126d1e909d294ec9a86","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-13T06:39:22Z","title_canon_sha256":"13ee7bd8a53f281f32b4eed8928919a7829518e08146142629a4abc9626e7c3e"},"schema_version":"1.0","source":{"id":"2502.09674","kind":"arxiv","version":4}},"canonical_sha256":"c28ff0966c8e0d08cf46b6fd75c14e976a9d2f44ac0244519608f327f0f796bd","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c28ff0966c8e0d08cf46b6fd75c14e976a9d2f44ac0244519608f327f0f796bd","first_computed_at":"2026-07-05T11:10:04.895155Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:10:04.895155Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"oMS/52DTuWk+ZYjAfU5P7GuoQSr2lxVX0Zh0r6ixf5Tu7h+skIxSTP6i0uhOUsi9PEEOfWVI5kt6dC+PYvqXCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T11:10:04.895690Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.09674","source_kind":"arxiv","source_version":4}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d9a01e0c74e3ab5d5427bb331cfc1aba5cd54da32d8860868acf4ac6ea87ba66","sha256:b323ec3099154d35923abc0edd9f88a96a0afbbab74d41bab5db5db75491e955"],"state_sha256":"a9b7378cb4c026af2b175e9448451960e92a9df4daa131c31d1137b06b118b08"}