{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:V7PQ6GUVO4KSM3HCYIGJZL3P4F","short_pith_number":"pith:V7PQ6GUV","schema_version":"1.0","canonical_sha256":"afdf0f1a957715266ce2c20c9caf6fe16f84db46dce1bf3e2ca0b0f64de3a531","source":{"kind":"arxiv","id":"2305.12132","version":2},"attestation_state":"computed","paper":{"title":"Can Public Large Language Models Help Private Cross-device Federated Learning?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Li, Boxin Wang, H. Brendan McMahan, Manzil Zaheer, Sewoong Oh, Yibo Jacky Zhang, Yuan Cao, Zheng Xu","submitted_at":"2023-05-20T07:55:58Z","abstract_excerpt":"We study (differentially) private federated learning (FL) of language models. The language models in cross-device FL are relatively small, which can be trained with meaningful formal user-level differential privacy (DP) guarantees when massive parallelism in training is enabled by the participation of a moderate size of users. Recently, public data has been used to improve privacy-utility trade-offs for both large and small language models. In this work, we provide a systematic study of using large-scale public data and LLMs to help differentially private training of on-device FL models, and f"},"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":"2305.12132","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2023-05-20T07:55:58Z","cross_cats_sorted":[],"title_canon_sha256":"49c3ba7d9019df1e14233affff6983286f744ca785a8596b5006f0e66565fffc","abstract_canon_sha256":"201d74f9137d87f2280b56659d82682df9dcc77a002b013654f57e9d4467c08c"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:07:27.512602Z","signature_b64":"rwRhkWh+D6c+KWfyjI2VPG7VmbfwP65ev/Y2jS1dfeH2wnlpHB+7IOB5t1zSOKXYLveBlb5o0pSeQfDg631VAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"afdf0f1a957715266ce2c20c9caf6fe16f84db46dce1bf3e2ca0b0f64de3a531","last_reissued_at":"2026-07-05T08:07:27.511904Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:07:27.511904Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Can Public Large Language Models Help Private Cross-device Federated Learning?","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Bo Li, Boxin Wang, H. Brendan McMahan, Manzil Zaheer, Sewoong Oh, Yibo Jacky Zhang, Yuan Cao, Zheng Xu","submitted_at":"2023-05-20T07:55:58Z","abstract_excerpt":"We study (differentially) private federated learning (FL) of language models. The language models in cross-device FL are relatively small, which can be trained with meaningful formal user-level differential privacy (DP) guarantees when massive parallelism in training is enabled by the participation of a moderate size of users. Recently, public data has been used to improve privacy-utility trade-offs for both large and small language models. In this work, we provide a systematic study of using large-scale public data and LLMs to help differentially private training of on-device FL models, and f"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2305.12132","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/2305.12132/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":"2305.12132","created_at":"2026-07-05T08:07:27.511999+00:00"},{"alias_kind":"arxiv_version","alias_value":"2305.12132v2","created_at":"2026-07-05T08:07:27.511999+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2305.12132","created_at":"2026-07-05T08:07:27.511999+00:00"},{"alias_kind":"pith_short_12","alias_value":"V7PQ6GUVO4KS","created_at":"2026-07-05T08:07:27.511999+00:00"},{"alias_kind":"pith_short_16","alias_value":"V7PQ6GUVO4KSM3HC","created_at":"2026-07-05T08:07:27.511999+00:00"},{"alias_kind":"pith_short_8","alias_value":"V7PQ6GUV","created_at":"2026-07-05T08:07:27.511999+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F","json":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F.json","graph_json":"https://pith.science/api/pith-number/V7PQ6GUVO4KSM3HCYIGJZL3P4F/graph.json","events_json":"https://pith.science/api/pith-number/V7PQ6GUVO4KSM3HCYIGJZL3P4F/events.json","paper":"https://pith.science/paper/V7PQ6GUV"},"agent_actions":{"view_html":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F","download_json":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F.json","view_paper":"https://pith.science/paper/V7PQ6GUV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2305.12132&json=true","fetch_graph":"https://pith.science/api/pith-number/V7PQ6GUVO4KSM3HCYIGJZL3P4F/graph.json","fetch_events":"https://pith.science/api/pith-number/V7PQ6GUVO4KSM3HCYIGJZL3P4F/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F/action/timestamp_anchor","attest_storage":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F/action/storage_attestation","attest_author":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F/action/author_attestation","sign_citation":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F/action/citation_signature","submit_replication":"https://pith.science/pith/V7PQ6GUVO4KSM3HCYIGJZL3P4F/action/replication_record"}},"created_at":"2026-07-05T08:07:27.511999+00:00","updated_at":"2026-07-05T08:07:27.511999+00:00"}