{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:ILHGCELOWALIZMJJO3RH2EFFLP","short_pith_number":"pith:ILHGCELO","schema_version":"1.0","canonical_sha256":"42ce61116eb0168cb12976e27d10a55bc1a3391501873e69ac7c1f956e51a3d6","source":{"kind":"arxiv","id":"2405.06368","version":4},"attestation_state":"computed","paper":{"title":"DP-DyLoRA: Fine-Tuning Transformer-Based Models On-Device under Differentially Private Federated Learning using Dynamic Low-Rank Adaptation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR","cs.DC"],"primary_cat":"cs.LG","authors_text":"David Tuckey, Haaris Mehmood, Jie Xu, Karthikeyan Saravanan, Mete Ozay, Rogier Van Dalen","submitted_at":"2024-05-10T10:10:37Z","abstract_excerpt":"Federated learning (FL) allows clients to collaboratively train a global model without sharing their local data with a server. However, clients' contributions to the server can still leak sensitive information. Differential privacy (DP) addresses such leakage by providing formal privacy guarantees, with mechanisms that add randomness to the clients' contributions. The randomness makes it infeasible to train large transformer-based models, common in modern federated learning systems. In this work, we empirically evaluate the practicality of fine-tuning large scale on-device transformer-based mo"},"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":"2405.06368","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2024-05-10T10:10:37Z","cross_cats_sorted":["cs.CR","cs.DC"],"title_canon_sha256":"37dd319ca59f7782c1110b0683efe46bc5169081f92eee44a41fb6ebf9b75e1e","abstract_canon_sha256":"37baf28524715279fa6d28d39ce764af5672ff378a45c723f8902a8361336edb"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:15:16.787764Z","signature_b64":"iHhfu/GbYOelogaYhupuZuN5NtPRwfJupT8lO5o76uP32Zb0lwZ+VGR5JBRyHlWh5rqQI0LcKq4ek4XkB40YAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"42ce61116eb0168cb12976e27d10a55bc1a3391501873e69ac7c1f956e51a3d6","last_reissued_at":"2026-07-05T10:15:16.787254Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:15:16.787254Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"DP-DyLoRA: Fine-Tuning Transformer-Based Models On-Device under Differentially Private Federated Learning using Dynamic Low-Rank Adaptation","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":["cs.CR","cs.DC"],"primary_cat":"cs.LG","authors_text":"David Tuckey, Haaris Mehmood, Jie Xu, Karthikeyan Saravanan, Mete Ozay, Rogier Van Dalen","submitted_at":"2024-05-10T10:10:37Z","abstract_excerpt":"Federated learning (FL) allows clients to collaboratively train a global model without sharing their local data with a server. However, clients' contributions to the server can still leak sensitive information. Differential privacy (DP) addresses such leakage by providing formal privacy guarantees, with mechanisms that add randomness to the clients' contributions. The randomness makes it infeasible to train large transformer-based models, common in modern federated learning systems. In this work, we empirically evaluate the practicality of fine-tuning large scale on-device transformer-based mo"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.06368","kind":"arxiv","version":4},"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/2405.06368/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":"2405.06368","created_at":"2026-07-05T10:15:16.787312+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.06368v4","created_at":"2026-07-05T10:15:16.787312+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.06368","created_at":"2026-07-05T10:15:16.787312+00:00"},{"alias_kind":"pith_short_12","alias_value":"ILHGCELOWALI","created_at":"2026-07-05T10:15:16.787312+00:00"},{"alias_kind":"pith_short_16","alias_value":"ILHGCELOWALIZMJJ","created_at":"2026-07-05T10:15:16.787312+00:00"},{"alias_kind":"pith_short_8","alias_value":"ILHGCELO","created_at":"2026-07-05T10:15:16.787312+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2506.05640","citing_title":"FedShield-LLM: A Secure and Scalable Federated Fine-Tuned Large Language Model","ref_index":42,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08443","citing_title":"Improving Parameter-Efficient Federated Learning with Differentially Private Refactorization","ref_index":30,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10272","citing_title":"DP-LAC: Lightweight Adaptive Clipping for Differentially Private Federated Fine-tuning of Language Models","ref_index":43,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20596","citing_title":"Differentially Private Clustered Federated Learning with Privacy-Preserving Initialization and Normality-Driven Aggregation","ref_index":31,"is_internal_anchor":false},{"citing_arxiv_id":"2604.21197","citing_title":"Toward Efficient Membership Inference Attacks against Federated Large Language Models: A Projection Residual Approach","ref_index":14,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP","json":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP.json","graph_json":"https://pith.science/api/pith-number/ILHGCELOWALIZMJJO3RH2EFFLP/graph.json","events_json":"https://pith.science/api/pith-number/ILHGCELOWALIZMJJO3RH2EFFLP/events.json","paper":"https://pith.science/paper/ILHGCELO"},"agent_actions":{"view_html":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP","download_json":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP.json","view_paper":"https://pith.science/paper/ILHGCELO","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.06368&json=true","fetch_graph":"https://pith.science/api/pith-number/ILHGCELOWALIZMJJO3RH2EFFLP/graph.json","fetch_events":"https://pith.science/api/pith-number/ILHGCELOWALIZMJJO3RH2EFFLP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP/action/storage_attestation","attest_author":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP/action/author_attestation","sign_citation":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP/action/citation_signature","submit_replication":"https://pith.science/pith/ILHGCELOWALIZMJJO3RH2EFFLP/action/replication_record"}},"created_at":"2026-07-05T10:15:16.787312+00:00","updated_at":"2026-07-05T10:15:16.787312+00:00"}