{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:KZUYCD6XCV6EPWTCRYYNN4NGGX","short_pith_number":"pith:KZUYCD6X","canonical_record":{"source":{"id":"2608.08383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-09T00:29:54Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3cd01677d1b411fe489369fa572567775253e31fe72f46fe5317c328ab5da264","abstract_canon_sha256":"f0c786f6ae1b20b46047383b8b24e9460e2420515f1aa372e520c77c4cab5493"},"schema_version":"1.0"},"canonical_sha256":"5669810fd7157c47da628e30d6f1a635e6695123c6752ab4771c8325cf1654f8","source":{"kind":"arxiv","id":"2608.08383","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.08383","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"arxiv_version","alias_value":"2608.08383v1","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08383","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"pith_short_12","alias_value":"KZUYCD6XCV6E","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"pith_short_16","alias_value":"KZUYCD6XCV6EPWTC","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"pith_short_8","alias_value":"KZUYCD6X","created_at":"2026-08-11T01:22:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:KZUYCD6XCV6EPWTCRYYNN4NGGX","target":"record","payload":{"canonical_record":{"source":{"id":"2608.08383","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-09T00:29:54Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"3cd01677d1b411fe489369fa572567775253e31fe72f46fe5317c328ab5da264","abstract_canon_sha256":"f0c786f6ae1b20b46047383b8b24e9460e2420515f1aa372e520c77c4cab5493"},"schema_version":"1.0"},"canonical_sha256":"5669810fd7157c47da628e30d6f1a635e6695123c6752ab4771c8325cf1654f8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-08-11T01:22:29.580492Z","signature_b64":"OUDNDxj0mC/nGBLIZXrTPb7EYJztZzriX87s/SSZavyoDHihOp1/x6eAVDcHJZmdPQG4/6EIP6pirnTfZuHBDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5669810fd7157c47da628e30d6f1a635e6695123c6752ab4771c8325cf1654f8","last_reissued_at":"2026-08-11T01:22:29.578054Z","signature_status":"signed_v1","first_computed_at":"2026-08-11T01:22:29.578054Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2608.08383","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-08-11T01:22:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HUKSMibGp/CxBP81RJad0RPm1hFMI3SCRVwUGRI64Pg4DATtQMWEJy3DuyMhdh2eOu6N2+UITN3SiKx1sE2pCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T21:33:16.753292Z"},"content_sha256":"d05c0241a840b0e2556c2189d31dda59239e00be57a93ddcd0e0649064dbe9e5","schema_version":"1.0","event_id":"sha256:d05c0241a840b0e2556c2189d31dda59239e00be57a93ddcd0e0649064dbe9e5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:KZUYCD6XCV6EPWTCRYYNN4NGGX","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Safety Cost of Steering Vectors Is Separable and Reducible","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Gjergji Kasneci, Yuxiao Li","submitted_at":"2026-08-09T00:29:54Z","abstract_excerpt":"Steering vectors are a lightweight tool for controlling LLM behavior. However, emerging evidence shows that steering vectors can unintentionally compromise a model's safety mechanisms and increase compliance with harmful requests, while no effective mitigation yet exists. In this work, we show that this safety degradation arises from a separable component in the vector that disrupts the model's safety mechanisms but contributes little to the steering objective. We identify and remove this safety-degrading component, formulating the task as a constrained optimization problem solved through prim"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08383","kind":"arxiv","version":1},"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/2608.08383/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":"2026-08-11T01:22:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IcwEthcs7eCkuAGiy4aOW9sxuotZC7IsjXPOtRRiWLNrVzWSnipgj/WGmsuLmCy1LcymfvlD7AQBWN7W0fbuBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-18T21:33:16.753849Z"},"content_sha256":"5a4781b7b9f751258070978e5b4eb2d4cceee7747fe6facbce0d89ca22fadd76","schema_version":"1.0","event_id":"sha256:5a4781b7b9f751258070978e5b4eb2d4cceee7747fe6facbce0d89ca22fadd76"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX/bundle.json","state_url":"https://pith.science/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX/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":"2026-08-18T21:33:16Z","links":{"resolver":"https://pith.science/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX","bundle":"https://pith.science/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX/bundle.json","state":"https://pith.science/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX/state.json","well_known_bundle":"https://pith.science/.well-known/pith/KZUYCD6XCV6EPWTCRYYNN4NGGX/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:KZUYCD6XCV6EPWTCRYYNN4NGGX","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":"f0c786f6ae1b20b46047383b8b24e9460e2420515f1aa372e520c77c4cab5493","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-09T00:29:54Z","title_canon_sha256":"3cd01677d1b411fe489369fa572567775253e31fe72f46fe5317c328ab5da264"},"schema_version":"1.0","source":{"id":"2608.08383","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2608.08383","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"arxiv_version","alias_value":"2608.08383v1","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2608.08383","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"pith_short_12","alias_value":"KZUYCD6XCV6E","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"pith_short_16","alias_value":"KZUYCD6XCV6EPWTC","created_at":"2026-08-11T01:22:29Z"},{"alias_kind":"pith_short_8","alias_value":"KZUYCD6X","created_at":"2026-08-11T01:22:29Z"}],"graph_snapshots":[{"event_id":"sha256:5a4781b7b9f751258070978e5b4eb2d4cceee7747fe6facbce0d89ca22fadd76","target":"graph","created_at":"2026-08-11T01:22:29Z","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/2608.08383/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Steering vectors are a lightweight tool for controlling LLM behavior. However, emerging evidence shows that steering vectors can unintentionally compromise a model's safety mechanisms and increase compliance with harmful requests, while no effective mitigation yet exists. In this work, we show that this safety degradation arises from a separable component in the vector that disrupts the model's safety mechanisms but contributes little to the steering objective. We identify and remove this safety-degrading component, formulating the task as a constrained optimization problem solved through prim","authors_text":"Gjergji Kasneci, Yuxiao Li","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-09T00:29:54Z","title":"Safety Cost of Steering Vectors Is Separable and Reducible"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2608.08383","kind":"arxiv","version":1},"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:d05c0241a840b0e2556c2189d31dda59239e00be57a93ddcd0e0649064dbe9e5","target":"record","created_at":"2026-08-11T01:22:29Z","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":"f0c786f6ae1b20b46047383b8b24e9460e2420515f1aa372e520c77c4cab5493","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2026-08-09T00:29:54Z","title_canon_sha256":"3cd01677d1b411fe489369fa572567775253e31fe72f46fe5317c328ab5da264"},"schema_version":"1.0","source":{"id":"2608.08383","kind":"arxiv","version":1}},"canonical_sha256":"5669810fd7157c47da628e30d6f1a635e6695123c6752ab4771c8325cf1654f8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5669810fd7157c47da628e30d6f1a635e6695123c6752ab4771c8325cf1654f8","first_computed_at":"2026-08-11T01:22:29.578054Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-08-11T01:22:29.578054Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"OUDNDxj0mC/nGBLIZXrTPb7EYJztZzriX87s/SSZavyoDHihOp1/x6eAVDcHJZmdPQG4/6EIP6pirnTfZuHBDQ==","signature_status":"signed_v1","signed_at":"2026-08-11T01:22:29.580492Z","signed_message":"canonical_sha256_bytes"},"source_id":"2608.08383","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:d05c0241a840b0e2556c2189d31dda59239e00be57a93ddcd0e0649064dbe9e5","sha256:5a4781b7b9f751258070978e5b4eb2d4cceee7747fe6facbce0d89ca22fadd76"],"state_sha256":"79e8a94ddecb6c0a681f98831fc7bdf157b2c7617d78a8c448cc771d89f3435c"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Ejg++2XVlZWvK51vqXAQE6DerqT1szf8Tk/C3+bxOSWDbbfubTcWFe473m3EpoTlZ5g1k5AEZY1JzpBCfv+ZBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-18T21:33:16.757451Z","bundle_sha256":"69df778901bea27d811b753f28637961d3e7fa72f1d3b6138c456a9d726ef601"}}