{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:TVMN2MT74OUNEFFE4XOMU3KBZX","short_pith_number":"pith:TVMN2MT7","schema_version":"1.0","canonical_sha256":"9d58dd327fe3a8d214a4e5dcca6d41cde0c4fa258bbbda8dbf0561429c770c33","source":{"kind":"arxiv","id":"2410.01463","version":3},"attestation_state":"computed","paper":{"title":"Selective Aggregation for Low-Rank Adaptation in Federated Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Feifei Wang, Huijie Fan, Liangqiong Qu, Pengxin Guo, Shuang Zeng, Yanran Wang","submitted_at":"2024-10-02T12:14:36Z","abstract_excerpt":"We investigate LoRA in federated learning through the lens of the asymmetry analysis of the learned $A$ and $B$ matrices. In doing so, we uncover that $A$ matrices are responsible for learning general knowledge, while $B$ matrices focus on capturing client-specific knowledge. Based on this finding, we introduce Federated Share-A Low-Rank Adaptation (FedSA-LoRA), which employs two low-rank trainable matrices $A$ and $B$ to model the weight update, but only $A$ matrices are shared with the server for aggregation. Moreover, we delve into the relationship between the learned $A$ and $B$ matrices i"},"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":"2410.01463","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-10-02T12:14:36Z","cross_cats_sorted":[],"title_canon_sha256":"b080a9dd4840a3ba0219d612be2dff0edc9d25148045f74c29b3db63ace40df7","abstract_canon_sha256":"576ca9fa3f53cd9f91b92448c6dba823edf830d9aed5d6adcdd1d295887c0466"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:36:29.424993Z","signature_b64":"rCGMZXhZJUn0JkiJMoSR/L5oM9zzYf1d0FnRgRNAiUdB0KIi+bzhW2oBt8WTaXD3lU7NzdXCj5/F0O6Azj8HBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9d58dd327fe3a8d214a4e5dcca6d41cde0c4fa258bbbda8dbf0561429c770c33","last_reissued_at":"2026-07-05T10:36:29.423929Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:36:29.423929Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Selective Aggregation for Low-Rank Adaptation in Federated Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Feifei Wang, Huijie Fan, Liangqiong Qu, Pengxin Guo, Shuang Zeng, Yanran Wang","submitted_at":"2024-10-02T12:14:36Z","abstract_excerpt":"We investigate LoRA in federated learning through the lens of the asymmetry analysis of the learned $A$ and $B$ matrices. In doing so, we uncover that $A$ matrices are responsible for learning general knowledge, while $B$ matrices focus on capturing client-specific knowledge. Based on this finding, we introduce Federated Share-A Low-Rank Adaptation (FedSA-LoRA), which employs two low-rank trainable matrices $A$ and $B$ to model the weight update, but only $A$ matrices are shared with the server for aggregation. Moreover, we delve into the relationship between the learned $A$ and $B$ matrices i"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.01463","kind":"arxiv","version":3},"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/2410.01463/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":"2410.01463","created_at":"2026-07-05T10:36:29.424094+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.01463v3","created_at":"2026-07-05T10:36:29.424094+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.01463","created_at":"2026-07-05T10:36:29.424094+00:00"},{"alias_kind":"pith_short_12","alias_value":"TVMN2MT74OUN","created_at":"2026-07-05T10:36:29.424094+00:00"},{"alias_kind":"pith_short_16","alias_value":"TVMN2MT74OUNEFFE","created_at":"2026-07-05T10:36:29.424094+00:00"},{"alias_kind":"pith_short_8","alias_value":"TVMN2MT7","created_at":"2026-07-05T10:36:29.424094+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":5,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.31208","citing_title":"Probing Memorization of Tabular In-Context Learning","ref_index":133,"is_internal_anchor":false},{"citing_arxiv_id":"2509.06984","citing_title":"FediLoRA: Practical Federated Fine-Tuning of Foundation Models Under Missing-Modality Constraints","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2602.19926","citing_title":"Rethinking LoRA for Privacy-Preserving Federated Learning in Large Models","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2604.00050","citing_title":"Task-Centric Personalized Federated Fine-Tuning of Language Models","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24103","citing_title":"Fed-DLoRA: Efficient Wireless Federated Learning with Dynamic Low-Rank Adaptation","ref_index":17,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX","json":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX.json","graph_json":"https://pith.science/api/pith-number/TVMN2MT74OUNEFFE4XOMU3KBZX/graph.json","events_json":"https://pith.science/api/pith-number/TVMN2MT74OUNEFFE4XOMU3KBZX/events.json","paper":"https://pith.science/paper/TVMN2MT7"},"agent_actions":{"view_html":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX","download_json":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX.json","view_paper":"https://pith.science/paper/TVMN2MT7","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.01463&json=true","fetch_graph":"https://pith.science/api/pith-number/TVMN2MT74OUNEFFE4XOMU3KBZX/graph.json","fetch_events":"https://pith.science/api/pith-number/TVMN2MT74OUNEFFE4XOMU3KBZX/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX/action/timestamp_anchor","attest_storage":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX/action/storage_attestation","attest_author":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX/action/author_attestation","sign_citation":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX/action/citation_signature","submit_replication":"https://pith.science/pith/TVMN2MT74OUNEFFE4XOMU3KBZX/action/replication_record"}},"created_at":"2026-07-05T10:36:29.424094+00:00","updated_at":"2026-07-05T10:36:29.424094+00:00"}