{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:PJYQ6GPBZJDAIQBJS5RBLVU276","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":"a64ec8fe99eb87559f967005658d6fec4593f3fedff86588bdb086cfb926f6c5","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-27T15:59:45Z","title_canon_sha256":"e78a12aef9399632df4d159c8bcc8e1d358782af3fea44b4c153913690d77d24"},"schema_version":"1.0","source":{"id":"2605.28664","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.28664","created_at":"2026-05-28T02:04:59Z"},{"alias_kind":"arxiv_version","alias_value":"2605.28664v1","created_at":"2026-05-28T02:04:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28664","created_at":"2026-05-28T02:04:59Z"},{"alias_kind":"pith_short_12","alias_value":"PJYQ6GPBZJDA","created_at":"2026-05-28T02:04:59Z"},{"alias_kind":"pith_short_16","alias_value":"PJYQ6GPBZJDAIQBJ","created_at":"2026-05-28T02:04:59Z"},{"alias_kind":"pith_short_8","alias_value":"PJYQ6GPB","created_at":"2026-05-28T02:04:59Z"}],"graph_snapshots":[{"event_id":"sha256:93b8a4484a01fb712ba96b153fc1e2cb65b56131f3f10740ced73e89575540d9","target":"graph","created_at":"2026-05-28T02:04:59Z","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/2605.28664/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Safety detection models require examples of HHH (Helpful, Harmless, Honest)-violating outputs for robust generalization, however such examples are scarce. Activation Steering (AS) has emerged as a data-efficient method for generating target-concept-aligned responses. We investigate whether AS can generate high-quality training datasets for downstream classifiers, a question that remains untested. We present a two-fold study with intrinsic and extrinsic evaluation across $4$ concepts $\\times\\,2$ models $\\times\\,4$ steering methods. Intrinsically, beyond the field-standard rubric of steering suc","authors_text":"Anna Rumshisky, Leman Akoglu, Tootiya Giyahchi, Veena Padmanabhan, Vijeta Deshpande","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-27T15:59:45Z","title":"Activation Steering for Synthetic Data Generation: The Role of Diversity in Downstream Safety Detection"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28664","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":"sha256:0fffd774fefb3a0981a0f86fd1b03cca51273d241c8d2a067e993f482004d4c1","target":"record","created_at":"2026-05-28T02:04:59Z","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":"a64ec8fe99eb87559f967005658d6fec4593f3fedff86588bdb086cfb926f6c5","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-27T15:59:45Z","title_canon_sha256":"e78a12aef9399632df4d159c8bcc8e1d358782af3fea44b4c153913690d77d24"},"schema_version":"1.0","source":{"id":"2605.28664","kind":"arxiv","version":1}},"canonical_sha256":"7a710f19e1ca46044029976215d69affbccb0d06ebef06821ab37dc8e3c1ebbc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7a710f19e1ca46044029976215d69affbccb0d06ebef06821ab37dc8e3c1ebbc","first_computed_at":"2026-05-28T02:04:59.252223Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-28T02:04:59.252223Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"4KcRfICEJp9etHmNMOWT3thucW+th32YEZ7XD8WPD8EiDVKBx5TSIHElzAjy/ckvjRNHUJkLhsIk9OvUuK0gDw==","signature_status":"signed_v1","signed_at":"2026-05-28T02:04:59.252672Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.28664","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:0fffd774fefb3a0981a0f86fd1b03cca51273d241c8d2a067e993f482004d4c1","sha256:93b8a4484a01fb712ba96b153fc1e2cb65b56131f3f10740ced73e89575540d9"],"state_sha256":"b6ee8ac07939215268f4bf5baa349fe523b92c5fe531115c1d782f8e15cbf367"}