{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2019:VUL2H3MGHNF2ECC5QDYBQWQXBS","short_pith_number":"pith:VUL2H3MG","schema_version":"1.0","canonical_sha256":"ad17a3ed863b4ba2085d80f0185a170c83231a50482465f72eff12adad349988","source":{"kind":"arxiv","id":"1910.06378","version":4},"attestation_state":"computed","paper":{"title":"SCAFFOLD: Stochastic Controlled Averaging for Federated Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Ananda Theertha Suresh, Mehryar Mohri, Sai Praneeth Karimireddy, Sashank J. Reddi, Satyen Kale, Sebastian U. Stich","submitted_at":"2019-10-14T18:49:20Z","abstract_excerpt":"Federated Averaging (FedAvg) has emerged as the algorithm of choice for federated learning due to its simplicity and low communication cost. However, in spite of recent research efforts, its performance is not fully understood. We obtain tight convergence rates for FedAvg and prove that it suffers from `client-drift' when the data is heterogeneous (non-iid), resulting in unstable and slow convergence.\n  As a solution, we propose a new algorithm (SCAFFOLD) which uses control variates (variance reduction) to correct for the `client-drift' in its local updates. We prove that SCAFFOLD requires sig"},"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":"1910.06378","kind":"arxiv","version":4},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-10-14T18:49:20Z","cross_cats_sorted":["cs.DC","math.OC","stat.ML"],"title_canon_sha256":"196d28eaf09108c8c5356f988468a977076d5c4d6aea168bc8c4a4f4649f548e","abstract_canon_sha256":"1b9280cfc60afaeb34575b759fbcafa343e1f580615f2c493346eb2fdde375a4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T02:30:27.218492Z","signature_b64":"Axso5eAMBgP26T+z2IiGjTkkcYMKjMT9A5DcRHXlfQN4Ayud0Zg0izla8jUkC3+zEAdPZ0OfUHN7V0ctyPy5Ag==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ad17a3ed863b4ba2085d80f0185a170c83231a50482465f72eff12adad349988","last_reissued_at":"2026-07-05T02:30:27.217940Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T02:30:27.217940Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SCAFFOLD: Stochastic Controlled Averaging for Federated Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.DC","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Ananda Theertha Suresh, Mehryar Mohri, Sai Praneeth Karimireddy, Sashank J. Reddi, Satyen Kale, Sebastian U. Stich","submitted_at":"2019-10-14T18:49:20Z","abstract_excerpt":"Federated Averaging (FedAvg) has emerged as the algorithm of choice for federated learning due to its simplicity and low communication cost. However, in spite of recent research efforts, its performance is not fully understood. We obtain tight convergence rates for FedAvg and prove that it suffers from `client-drift' when the data is heterogeneous (non-iid), resulting in unstable and slow convergence.\n  As a solution, we propose a new algorithm (SCAFFOLD) which uses control variates (variance reduction) to correct for the `client-drift' in its local updates. We prove that SCAFFOLD requires sig"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1910.06378","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/1910.06378/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":"1910.06378","created_at":"2026-07-05T02:30:27.218006+00:00"},{"alias_kind":"arxiv_version","alias_value":"1910.06378v4","created_at":"2026-07-05T02:30:27.218006+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1910.06378","created_at":"2026-07-05T02:30:27.218006+00:00"},{"alias_kind":"pith_short_12","alias_value":"VUL2H3MGHNF2","created_at":"2026-07-05T02:30:27.218006+00:00"},{"alias_kind":"pith_short_16","alias_value":"VUL2H3MGHNF2ECC5","created_at":"2026-07-05T02:30:27.218006+00:00"},{"alias_kind":"pith_short_8","alias_value":"VUL2H3MG","created_at":"2026-07-05T02:30:27.218006+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":10,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.26037","citing_title":"FedReLa: Imbalanced Federated Learning via Re-Labeling","ref_index":1,"is_internal_anchor":false},{"citing_arxiv_id":"2607.00173","citing_title":"TallyTrain: Communication-Efficient Federated Distillation","ref_index":29,"is_internal_anchor":false},{"citing_arxiv_id":"2606.27622","citing_title":"FoggyTrust: Robust Federated Learning with Hierarchical Trust Networks","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30930","citing_title":"SGD at the Edge of Stability: Stochastic Stabilization with Large Learning Rates","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27892","citing_title":"FedEHR-Gen: Federated Synthetic Time-Series EHR Generation via Latent Space Alignment and Distribution-Aware Aggregation","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2501.06332","citing_title":"Aggregating Low Rank Adapters in Federated Fine-tuning","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2003.00295","citing_title":"Adaptive Federated Optimization","ref_index":220,"is_internal_anchor":false},{"citing_arxiv_id":"2605.13434","citing_title":"Rescaled Asynchronous SGD: Optimal Distributed Optimization under Data and System Heterogeneity","ref_index":50,"is_internal_anchor":false},{"citing_arxiv_id":"2604.25467","citing_title":"Subspace Optimization for Efficient Federated Learning under Heterogeneous Data","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.07888","citing_title":"Enhancing Federated Quadruplet Learning: Stochastic Client Selection and Embedding Stability Analysis","ref_index":12,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS","json":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS.json","graph_json":"https://pith.science/api/pith-number/VUL2H3MGHNF2ECC5QDYBQWQXBS/graph.json","events_json":"https://pith.science/api/pith-number/VUL2H3MGHNF2ECC5QDYBQWQXBS/events.json","paper":"https://pith.science/paper/VUL2H3MG"},"agent_actions":{"view_html":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS","download_json":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS.json","view_paper":"https://pith.science/paper/VUL2H3MG","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=1910.06378&json=true","fetch_graph":"https://pith.science/api/pith-number/VUL2H3MGHNF2ECC5QDYBQWQXBS/graph.json","fetch_events":"https://pith.science/api/pith-number/VUL2H3MGHNF2ECC5QDYBQWQXBS/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS/action/storage_attestation","attest_author":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS/action/author_attestation","sign_citation":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS/action/citation_signature","submit_replication":"https://pith.science/pith/VUL2H3MGHNF2ECC5QDYBQWQXBS/action/replication_record"}},"created_at":"2026-07-05T02:30:27.218006+00:00","updated_at":"2026-07-05T02:30:27.218006+00:00"}