{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:CCFHPOTSAETUW4PC7X53PFKPQP","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":"d67f737a0814f287f80ed66d38622efa3faf1737c10a100dfb2acd4e3000e9dd","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-30T17:58:36Z","title_canon_sha256":"d023869629fcf99ad66adffb275baff89c03fa5d15d6d3ceba820d08c8407df9"},"schema_version":"1.0","source":{"id":"2501.18532","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2501.18532","created_at":"2026-07-05T10:35:45Z"},{"alias_kind":"arxiv_version","alias_value":"2501.18532v2","created_at":"2026-07-05T10:35:45Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2501.18532","created_at":"2026-07-05T10:35:45Z"},{"alias_kind":"pith_short_12","alias_value":"CCFHPOTSAETU","created_at":"2026-07-05T10:35:45Z"},{"alias_kind":"pith_short_16","alias_value":"CCFHPOTSAETUW4PC","created_at":"2026-07-05T10:35:45Z"},{"alias_kind":"pith_short_8","alias_value":"CCFHPOTS","created_at":"2026-07-05T10:35:45Z"}],"graph_snapshots":[{"event_id":"sha256:2a2726948cd2e2a4476c90e9c8213597397c248a15d5387945badc33e60756cb","target":"graph","created_at":"2026-07-05T10:35:45Z","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/2501.18532/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Aligning Large Language Models (LLMs) with human values and away from undesirable behaviors (such as hallucination) has become increasingly important. Recently, steering LLMs towards a desired behavior via activation editing has emerged as an effective method to mitigate harmful generations at inference-time. Activation editing modifies LLM representations by preserving information from positive demonstrations (e.g., truthful) and minimising information from negative demonstrations (e.g., hallucinations). When these demonstrations come from a private dataset, the aligned LLM may leak private i","authors_text":"Amartya Sanyal, Anmol Goel, Iryna Gurevych, Yaxi Hu","cross_cats":["cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-30T17:58:36Z","title":"Differentially Private Steering for Large Language Model Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2501.18532","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":"sha256:263620a53e53173c11971942998dc8405e8255e90e3b0eb8b8b7db1d7b2aa67b","target":"record","created_at":"2026-07-05T10:35:45Z","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":"d67f737a0814f287f80ed66d38622efa3faf1737c10a100dfb2acd4e3000e9dd","cross_cats_sorted":["cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-01-30T17:58:36Z","title_canon_sha256":"d023869629fcf99ad66adffb275baff89c03fa5d15d6d3ceba820d08c8407df9"},"schema_version":"1.0","source":{"id":"2501.18532","kind":"arxiv","version":2}},"canonical_sha256":"108a77ba7201274b71e2fdfbb7954f83c046298f7b4cf02d1baafbf43546411e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"108a77ba7201274b71e2fdfbb7954f83c046298f7b4cf02d1baafbf43546411e","first_computed_at":"2026-07-05T10:35:45.028722Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:35:45.028722Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"bRzYGdc4V2tMA/4qvDRYserkjZ2ErVth9Ud05LjmslzvTyV9kjSfZIY8kUSU4s21ZxqCd+KyXUzJr7K23KTICg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:35:45.029313Z","signed_message":"canonical_sha256_bytes"},"source_id":"2501.18532","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:263620a53e53173c11971942998dc8405e8255e90e3b0eb8b8b7db1d7b2aa67b","sha256:2a2726948cd2e2a4476c90e9c8213597397c248a15d5387945badc33e60756cb"],"state_sha256":"10bdcb865e936de84d568457fc0692cfb79f27b08e34b2bd4b92e6554268517e"}