{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:HQQ2IAHJRFPIWBCB6J4O2JOM7O","short_pith_number":"pith:HQQ2IAHJ","schema_version":"1.0","canonical_sha256":"3c21a400e9895e8b0441f278ed25ccfb89ff03f67b3e5f1e6d17d28c30c551f4","source":{"kind":"arxiv","id":"2310.02224","version":1},"attestation_state":"computed","paper":{"title":"Can Language Models be Instructed to Protect Personal Information?","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alan Ritter, Ethan Mendes, Sauvik Das, Wei Xu, Yang Chen","submitted_at":"2023-10-03T17:30:33Z","abstract_excerpt":"Large multimodal language models have proven transformative in numerous applications. However, these models have been shown to memorize and leak pre-training data, raising serious user privacy and information security concerns. While data leaks should be prevented, it is also crucial to examine the trade-off between the privacy protection and model utility of proposed approaches. In this paper, we introduce PrivQA -- a multimodal benchmark to assess this privacy/utility trade-off when a model is instructed to protect specific categories of personal information in a simulated scenario. We also "},"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":"2310.02224","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2023-10-03T17:30:33Z","cross_cats_sorted":[],"title_canon_sha256":"3e275dbff2cc5efcd8b34d490cf26f2f6a7729bf1576aa443ea3b58b6c2a450a","abstract_canon_sha256":"5cc06a4db670b0955773e2919a42afab85d530bf85f2b1744e09a7c880639182"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:56:51.534381Z","signature_b64":"sgLB3kK8P6nSqIkDIlFE6Pto3A049Y/zHW9E2XRt30lpxW7uEejP257F9prZ2y3cd0YJgSVQuE8K5XUaVW/9Bg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3c21a400e9895e8b0441f278ed25ccfb89ff03f67b3e5f1e6d17d28c30c551f4","last_reissued_at":"2026-07-05T06:56:51.533877Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:56:51.533877Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can Language Models be Instructed to Protect Personal Information?","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alan Ritter, Ethan Mendes, Sauvik Das, Wei Xu, Yang Chen","submitted_at":"2023-10-03T17:30:33Z","abstract_excerpt":"Large multimodal language models have proven transformative in numerous applications. However, these models have been shown to memorize and leak pre-training data, raising serious user privacy and information security concerns. While data leaks should be prevented, it is also crucial to examine the trade-off between the privacy protection and model utility of proposed approaches. In this paper, we introduce PrivQA -- a multimodal benchmark to assess this privacy/utility trade-off when a model is instructed to protect specific categories of personal information in a simulated scenario. We also "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.02224","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/2310.02224/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":"2310.02224","created_at":"2026-07-05T06:56:51.533944+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.02224v1","created_at":"2026-07-05T06:56:51.533944+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.02224","created_at":"2026-07-05T06:56:51.533944+00:00"},{"alias_kind":"pith_short_12","alias_value":"HQQ2IAHJRFPI","created_at":"2026-07-05T06:56:51.533944+00:00"},{"alias_kind":"pith_short_16","alias_value":"HQQ2IAHJRFPIWBCB","created_at":"2026-07-05T06:56:51.533944+00:00"},{"alias_kind":"pith_short_8","alias_value":"HQQ2IAHJ","created_at":"2026-07-05T06:56:51.533944+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2507.10695","citing_title":"Exploring User Security and Privacy Attitudes and Concerns Toward the Use of General-Purpose LLM Chatbots for Mental Health","ref_index":21,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O","json":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O.json","graph_json":"https://pith.science/api/pith-number/HQQ2IAHJRFPIWBCB6J4O2JOM7O/graph.json","events_json":"https://pith.science/api/pith-number/HQQ2IAHJRFPIWBCB6J4O2JOM7O/events.json","paper":"https://pith.science/paper/HQQ2IAHJ"},"agent_actions":{"view_html":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O","download_json":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O.json","view_paper":"https://pith.science/paper/HQQ2IAHJ","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.02224&json=true","fetch_graph":"https://pith.science/api/pith-number/HQQ2IAHJRFPIWBCB6J4O2JOM7O/graph.json","fetch_events":"https://pith.science/api/pith-number/HQQ2IAHJRFPIWBCB6J4O2JOM7O/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O/action/timestamp_anchor","attest_storage":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O/action/storage_attestation","attest_author":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O/action/author_attestation","sign_citation":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O/action/citation_signature","submit_replication":"https://pith.science/pith/HQQ2IAHJRFPIWBCB6J4O2JOM7O/action/replication_record"}},"created_at":"2026-07-05T06:56:51.533944+00:00","updated_at":"2026-07-05T06:56:51.533944+00:00"}