{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:AHUFTKNSH6TLBL7IVLRYRP4YKK","short_pith_number":"pith:AHUFTKNS","schema_version":"1.0","canonical_sha256":"01e859a9b23fa6b0afe8aae388bf9852ade519d095675ee1cc47d3364989e04b","source":{"kind":"arxiv","id":"2502.15680","version":2},"attestation_state":"computed","paper":{"title":"Privacy Ripple Effects from Adding or Removing Personal Information in Language Model Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.CL","authors_text":"Christopher A. Choquette-Choo, David A. Smith, Jaydeep Borkar, Katherine Lee, Matthew Jagielski, Niloofar Mireshghallah","submitted_at":"2025-02-21T18:59:14Z","abstract_excerpt":"Due to the sensitive nature of personally identifiable information (PII), its owners may have the authority to control its inclusion or request its removal from large-language model (LLM) training. Beyond this, PII may be added or removed from training datasets due to evolving dataset curation techniques, because they were newly scraped for retraining, or because they were included in a new downstream fine-tuning stage. We find that the amount and ease of PII memorization is a dynamic property of a model that evolves throughout training pipelines and depends on commonly altered design choices."},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2502.15680","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-02-21T18:59:14Z","cross_cats_sorted":["cs.CR"],"title_canon_sha256":"6ea5b144b4e139afcc176fe964b3f5193bb28a77f998c7c61ad3f9e6100bbcec","abstract_canon_sha256":"40606d2440b540e3cd3def70b40fa6598446f40389760e09dda65a801a2c9e76"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:27:24.288483Z","signature_b64":"c+A2EC8R76Zyzlao1eWEZoRC1ujm0/N7hls016t5sVD6Ynwl8QbjM1aVWKat3LqhUqV3j2JKDI4qJB+vSNkDBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"01e859a9b23fa6b0afe8aae388bf9852ade519d095675ee1cc47d3364989e04b","last_reissued_at":"2026-07-05T11:27:24.287779Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:27:24.287779Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Privacy Ripple Effects from Adding or Removing Personal Information in Language Model Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CR"],"primary_cat":"cs.CL","authors_text":"Christopher A. Choquette-Choo, David A. Smith, Jaydeep Borkar, Katherine Lee, Matthew Jagielski, Niloofar Mireshghallah","submitted_at":"2025-02-21T18:59:14Z","abstract_excerpt":"Due to the sensitive nature of personally identifiable information (PII), its owners may have the authority to control its inclusion or request its removal from large-language model (LLM) training. Beyond this, PII may be added or removed from training datasets due to evolving dataset curation techniques, because they were newly scraped for retraining, or because they were included in a new downstream fine-tuning stage. We find that the amount and ease of PII memorization is a dynamic property of a model that evolves throughout training pipelines and depends on commonly altered design choices."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.15680","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/2502.15680/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2502.15680","created_at":"2026-07-05T11:27:24.287871+00:00"},{"alias_kind":"arxiv_version","alias_value":"2502.15680v2","created_at":"2026-07-05T11:27:24.287871+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.15680","created_at":"2026-07-05T11:27:24.287871+00:00"},{"alias_kind":"pith_short_12","alias_value":"AHUFTKNSH6TL","created_at":"2026-07-05T11:27:24.287871+00:00"},{"alias_kind":"pith_short_16","alias_value":"AHUFTKNSH6TLBL7I","created_at":"2026-07-05T11:27:24.287871+00:00"},{"alias_kind":"pith_short_8","alias_value":"AHUFTKNS","created_at":"2026-07-05T11:27:24.287871+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.17185","citing_title":"A Common Pool of Privacy Problems: Legal and Technical Lessons from a Large-Scale Web-Scraped Machine Learning Dataset","ref_index":20,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK","json":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK.json","graph_json":"https://pith.science/api/pith-number/AHUFTKNSH6TLBL7IVLRYRP4YKK/graph.json","events_json":"https://pith.science/api/pith-number/AHUFTKNSH6TLBL7IVLRYRP4YKK/events.json","paper":"https://pith.science/paper/AHUFTKNS"},"agent_actions":{"view_html":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK","download_json":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK.json","view_paper":"https://pith.science/paper/AHUFTKNS","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2502.15680&json=true","fetch_graph":"https://pith.science/api/pith-number/AHUFTKNSH6TLBL7IVLRYRP4YKK/graph.json","fetch_events":"https://pith.science/api/pith-number/AHUFTKNSH6TLBL7IVLRYRP4YKK/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK/action/timestamp_anchor","attest_storage":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK/action/storage_attestation","attest_author":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK/action/author_attestation","sign_citation":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK/action/citation_signature","submit_replication":"https://pith.science/pith/AHUFTKNSH6TLBL7IVLRYRP4YKK/action/replication_record"}},"created_at":"2026-07-05T11:27:24.287871+00:00","updated_at":"2026-07-05T11:27:24.287871+00:00"}