{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:IXXD4D3VL5QYBZAPHJDSARJ47O","short_pith_number":"pith:IXXD4D3V","canonical_record":{"source":{"id":"2411.02193","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T15:46:20Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"0c9447467dc1a723e83580e0370f954cd71cbbc1f8f9f4fe4dec0c42d5d8757a","abstract_canon_sha256":"24486a5506058d149ff28046d4847b0f833373cda1c6a177d9cae6486e2ef69e"},"schema_version":"1.0"},"canonical_sha256":"45ee3e0f755f6180e40f3a4720453cfb883644640a19f8b03f074e981414c4fb","source":{"kind":"arxiv","id":"2411.02193","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.02193","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"arxiv_version","alias_value":"2411.02193v2","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02193","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"pith_short_12","alias_value":"IXXD4D3VL5QY","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"pith_short_16","alias_value":"IXXD4D3VL5QYBZAP","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"pith_short_8","alias_value":"IXXD4D3V","created_at":"2026-07-05T09:38:31Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:IXXD4D3VL5QYBZAPHJDSARJ47O","target":"record","payload":{"canonical_record":{"source":{"id":"2411.02193","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T15:46:20Z","cross_cats_sorted":["cs.AI","cs.CL"],"title_canon_sha256":"0c9447467dc1a723e83580e0370f954cd71cbbc1f8f9f4fe4dec0c42d5d8757a","abstract_canon_sha256":"24486a5506058d149ff28046d4847b0f833373cda1c6a177d9cae6486e2ef69e"},"schema_version":"1.0"},"canonical_sha256":"45ee3e0f755f6180e40f3a4720453cfb883644640a19f8b03f074e981414c4fb","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:38:31.870916Z","signature_b64":"o8FiK560p1OaU9HqWyIphb1Y80n+T9S3xazkf72uBnQYfE0/TbWQ73jNXjr+p8xVmOJ5n2mTWCnULnNGvNPtDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"45ee3e0f755f6180e40f3a4720453cfb883644640a19f8b03f074e981414c4fb","last_reissued_at":"2026-07-05T09:38:31.870460Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:38:31.870460Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2411.02193","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":"2026-07-05T09:38:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9D3A+uO4Bk/ZC2rRzY84qofm9M88wvAIhaZ86HrdSF7rUCErwRLVSvoUdv7ZLebsziPiUtTQTqaQNjQOTEk7CQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T01:30:37.713767Z"},"content_sha256":"f057fb602f14d90d541db747931cb7ac84245cbb3c3919d45b69c7e141a34eae","schema_version":"1.0","event_id":"sha256:f057fb602f14d90d541db747931cb7ac84245cbb3c3919d45b69c7e141a34eae"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:IXXD4D3VL5QYBZAPHJDSARJ47O","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Improving Steering Vectors by Targeting Sparse Autoencoder Features","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CL"],"primary_cat":"cs.LG","authors_text":"Arthur Conmy, Matthew Siu, Sviatoslav Chalnev","submitted_at":"2024-11-04T15:46:20Z","abstract_excerpt":"To control the behavior of language models, steering methods attempt to ensure that outputs of the model satisfy specific pre-defined properties. Adding steering vectors to the model is a promising method of model control that is easier than finetuning, and may be more robust than prompting. However, it can be difficult to anticipate the effects of steering vectors produced by methods such as CAA [Panickssery et al., 2024] or the direct use of SAE latents [Templeton et al., 2024]. In our work, we address this issue by using SAEs to measure the effects of steering vectors, giving us a method th"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02193","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/2411.02193/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":"2026-07-05T09:38:31Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"jHkucwu3aTJpOAhyWzccsH1inzODIDjcfgcK2ci8E5eXWLzYxYUIKEPB6N/sMrtvo6wdRSK530dn6mkTbrv0DA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T01:30:37.714293Z"},"content_sha256":"defa659b97971804bddceb7abd3d0fbe41a1bfaae000d9677109a70b26aa240f","schema_version":"1.0","event_id":"sha256:defa659b97971804bddceb7abd3d0fbe41a1bfaae000d9677109a70b26aa240f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/IXXD4D3VL5QYBZAPHJDSARJ47O/bundle.json","state_url":"https://pith.science/pith/IXXD4D3VL5QYBZAPHJDSARJ47O/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/IXXD4D3VL5QYBZAPHJDSARJ47O/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":"2026-08-07T01:30:37Z","links":{"resolver":"https://pith.science/pith/IXXD4D3VL5QYBZAPHJDSARJ47O","bundle":"https://pith.science/pith/IXXD4D3VL5QYBZAPHJDSARJ47O/bundle.json","state":"https://pith.science/pith/IXXD4D3VL5QYBZAPHJDSARJ47O/state.json","well_known_bundle":"https://pith.science/.well-known/pith/IXXD4D3VL5QYBZAPHJDSARJ47O/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:IXXD4D3VL5QYBZAPHJDSARJ47O","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":"24486a5506058d149ff28046d4847b0f833373cda1c6a177d9cae6486e2ef69e","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T15:46:20Z","title_canon_sha256":"0c9447467dc1a723e83580e0370f954cd71cbbc1f8f9f4fe4dec0c42d5d8757a"},"schema_version":"1.0","source":{"id":"2411.02193","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2411.02193","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"arxiv_version","alias_value":"2411.02193v2","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.02193","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"pith_short_12","alias_value":"IXXD4D3VL5QY","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"pith_short_16","alias_value":"IXXD4D3VL5QYBZAP","created_at":"2026-07-05T09:38:31Z"},{"alias_kind":"pith_short_8","alias_value":"IXXD4D3V","created_at":"2026-07-05T09:38:31Z"}],"graph_snapshots":[{"event_id":"sha256:defa659b97971804bddceb7abd3d0fbe41a1bfaae000d9677109a70b26aa240f","target":"graph","created_at":"2026-07-05T09:38:31Z","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/2411.02193/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"To control the behavior of language models, steering methods attempt to ensure that outputs of the model satisfy specific pre-defined properties. Adding steering vectors to the model is a promising method of model control that is easier than finetuning, and may be more robust than prompting. However, it can be difficult to anticipate the effects of steering vectors produced by methods such as CAA [Panickssery et al., 2024] or the direct use of SAE latents [Templeton et al., 2024]. In our work, we address this issue by using SAEs to measure the effects of steering vectors, giving us a method th","authors_text":"Arthur Conmy, Matthew Siu, Sviatoslav Chalnev","cross_cats":["cs.AI","cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T15:46:20Z","title":"Improving Steering Vectors by Targeting Sparse Autoencoder Features"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.02193","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:f057fb602f14d90d541db747931cb7ac84245cbb3c3919d45b69c7e141a34eae","target":"record","created_at":"2026-07-05T09:38:31Z","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":"24486a5506058d149ff28046d4847b0f833373cda1c6a177d9cae6486e2ef69e","cross_cats_sorted":["cs.AI","cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-04T15:46:20Z","title_canon_sha256":"0c9447467dc1a723e83580e0370f954cd71cbbc1f8f9f4fe4dec0c42d5d8757a"},"schema_version":"1.0","source":{"id":"2411.02193","kind":"arxiv","version":2}},"canonical_sha256":"45ee3e0f755f6180e40f3a4720453cfb883644640a19f8b03f074e981414c4fb","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"45ee3e0f755f6180e40f3a4720453cfb883644640a19f8b03f074e981414c4fb","first_computed_at":"2026-07-05T09:38:31.870460Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T09:38:31.870460Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"o8FiK560p1OaU9HqWyIphb1Y80n+T9S3xazkf72uBnQYfE0/TbWQ73jNXjr+p8xVmOJ5n2mTWCnULnNGvNPtDg==","signature_status":"signed_v1","signed_at":"2026-07-05T09:38:31.870916Z","signed_message":"canonical_sha256_bytes"},"source_id":"2411.02193","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:f057fb602f14d90d541db747931cb7ac84245cbb3c3919d45b69c7e141a34eae","sha256:defa659b97971804bddceb7abd3d0fbe41a1bfaae000d9677109a70b26aa240f"],"state_sha256":"4cf189539101dc7133ba24f9d6ddbe09cb1013a6342950797c4c1ed2c6c78a4f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"z4BNp7BzDdApnyAxoKU25/8EqU1HNua2uhl7sr08veEyEm9BIZogsLWJOd2OzGX9WK/BHzETKDntmgIHIZL+Bw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T01:30:37.719053Z","bundle_sha256":"545e7dafdb62e26a907fa175752b709485857a345f223fb75460465084bf49fc"}}