{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:RN2WUNELUQVPN72B6PEPTI3F4D","short_pith_number":"pith:RN2WUNEL","schema_version":"1.0","canonical_sha256":"8b756a348ba42af6ff41f3c8f9a365e0e4be15356f0991915882a0dcda38d6e7","source":{"kind":"arxiv","id":"2205.10162","version":2},"attestation_state":"computed","paper":{"title":"FedAdapter: Efficient Federated Learning for Modern NLP","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dongqi Cai, Felix Xiaozhu Lin, Mengwei Xu, Shangguang Wang, Yaozong Wu","submitted_at":"2022-05-20T13:10:43Z","abstract_excerpt":"Transformer-based pre-trained models have revolutionized NLP for superior performance and generality. Fine-tuning pre-trained models for downstream tasks often requires private data, for which federated learning is the de-facto approach (i.e., FedNLP). However, our measurements show that FedNLP is prohibitively slow due to the large model sizes and the resultant high network/computation cost. Towards practical FedNLP, we identify as the key building blocks adapters, small bottleneck modules inserted at a variety of model layers. A key challenge is to properly configure the depth and width of a"},"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":"2205.10162","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2022-05-20T13:10:43Z","cross_cats_sorted":[],"title_canon_sha256":"6522ee2ab2f365721e816e3ae95893253c3ef9afcc19a48154b34600ec0749ac","abstract_canon_sha256":"2879d9b3e932f4894e7d88e99ceb11d0e22358c76e1d53e85ef8f7910a8305d5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:08:04.985282Z","signature_b64":"V6ryOt7tO0folZnQyN4fFq41ITWI009XwVZR/b8xILBbIMzMrkWXgPU+7DWGpTtc6N5EEzssOUlx6r+pyfVsAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"8b756a348ba42af6ff41f3c8f9a365e0e4be15356f0991915882a0dcda38d6e7","last_reissued_at":"2026-07-05T06:08:04.984792Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:08:04.984792Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"FedAdapter: Efficient Federated Learning for Modern NLP","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Dongqi Cai, Felix Xiaozhu Lin, Mengwei Xu, Shangguang Wang, Yaozong Wu","submitted_at":"2022-05-20T13:10:43Z","abstract_excerpt":"Transformer-based pre-trained models have revolutionized NLP for superior performance and generality. Fine-tuning pre-trained models for downstream tasks often requires private data, for which federated learning is the de-facto approach (i.e., FedNLP). However, our measurements show that FedNLP is prohibitively slow due to the large model sizes and the resultant high network/computation cost. Towards practical FedNLP, we identify as the key building blocks adapters, small bottleneck modules inserted at a variety of model layers. A key challenge is to properly configure the depth and width of a"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.10162","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/2205.10162/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":"2205.10162","created_at":"2026-07-05T06:08:04.984853+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.10162v2","created_at":"2026-07-05T06:08:04.984853+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.10162","created_at":"2026-07-05T06:08:04.984853+00:00"},{"alias_kind":"pith_short_12","alias_value":"RN2WUNELUQVP","created_at":"2026-07-05T06:08:04.984853+00:00"},{"alias_kind":"pith_short_16","alias_value":"RN2WUNELUQVPN72B","created_at":"2026-07-05T06:08:04.984853+00:00"},{"alias_kind":"pith_short_8","alias_value":"RN2WUNEL","created_at":"2026-07-05T06:08:04.984853+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":4,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2411.11707","citing_title":"Federated Co-tuning Framework for Large and Small Language Models","ref_index":3,"is_internal_anchor":false},{"citing_arxiv_id":"2604.19015","citing_title":"FedProxy: Federated Fine-Tuning of LLMs via Proxy SLMs and Heterogeneity-Aware Fusion","ref_index":39,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06819","citing_title":"Beyond End-to-End: Dynamic Chain Optimization for Private LLM Adaptation on the Edge","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.06297","citing_title":"FedSpy-LLM: Towards Scalable and Generalizable Data Reconstruction Attacks from Gradients on LLMs","ref_index":22,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D","json":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D.json","graph_json":"https://pith.science/api/pith-number/RN2WUNELUQVPN72B6PEPTI3F4D/graph.json","events_json":"https://pith.science/api/pith-number/RN2WUNELUQVPN72B6PEPTI3F4D/events.json","paper":"https://pith.science/paper/RN2WUNEL"},"agent_actions":{"view_html":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D","download_json":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D.json","view_paper":"https://pith.science/paper/RN2WUNEL","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.10162&json=true","fetch_graph":"https://pith.science/api/pith-number/RN2WUNELUQVPN72B6PEPTI3F4D/graph.json","fetch_events":"https://pith.science/api/pith-number/RN2WUNELUQVPN72B6PEPTI3F4D/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D/action/timestamp_anchor","attest_storage":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D/action/storage_attestation","attest_author":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D/action/author_attestation","sign_citation":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D/action/citation_signature","submit_replication":"https://pith.science/pith/RN2WUNELUQVPN72B6PEPTI3F4D/action/replication_record"}},"created_at":"2026-07-05T06:08:04.984853+00:00","updated_at":"2026-07-05T06:08:04.984853+00:00"}