{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:UQ6G335HDMWH5OQA2UQXOUBOKO","short_pith_number":"pith:UQ6G335H","schema_version":"1.0","canonical_sha256":"a43c6defa71b2c7eba00d52177502e53ac94606820c17a517ba6df78ff9db160","source":{"kind":"arxiv","id":"2601.20738","version":1},"attestation_state":"computed","paper":{"title":"SA-PEF: Step-Ahead Partial Error Feedback for Efficient Federated Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC","eess.SP","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Dawit Kiros Redie, Reza Arablouei, Stefan Werner","submitted_at":"2026-01-28T16:10:49Z","abstract_excerpt":"Biased gradient compression with error feedback (EF) reduces communication in federated learning (FL), but under non-IID data, the residual error can decay slowly, causing gradient mismatch and stalled progress in the early rounds. We propose step-ahead partial error feedback (SA-PEF), which integrates step-ahead (SA) correction with partial error feedback (PEF). SA-PEF recovers EF when the step-ahead coefficient $\\alpha=0$ and step-ahead EF (SAEF) when $\\alpha=1$. For non-convex objectives and $\\delta$-contractive compressors, we establish a second-moment bound and a residual recursion that g"},"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":"2601.20738","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-01-28T16:10:49Z","cross_cats_sorted":["cs.DC","eess.SP","math.OC","stat.ML"],"title_canon_sha256":"1b1e11616384a895dcbe1f61e4e8f5be863e23f1b8dac29db2890795cf5dbc77","abstract_canon_sha256":"a2b343022f21a3aa016815ccbab502c9e138ac4489a3f068178d3079254aaba6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-26T01:03:23.383540Z","signature_b64":"PMhGXRE4Ynluh/p73OrELxBYCrLn8ZbtssEDEiqr2mvC6MwjXM/ZZD904N4r/oSbXnkSr1fwS8pZTfz2+V87Dg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a43c6defa71b2c7eba00d52177502e53ac94606820c17a517ba6df78ff9db160","last_reissued_at":"2026-05-26T01:03:23.382549Z","signature_status":"signed_v1","first_computed_at":"2026-05-26T01:03:23.382549Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"SA-PEF: Step-Ahead Partial Error Feedback for Efficient Federated Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.DC","eess.SP","math.OC","stat.ML"],"primary_cat":"cs.LG","authors_text":"Dawit Kiros Redie, Reza Arablouei, Stefan Werner","submitted_at":"2026-01-28T16:10:49Z","abstract_excerpt":"Biased gradient compression with error feedback (EF) reduces communication in federated learning (FL), but under non-IID data, the residual error can decay slowly, causing gradient mismatch and stalled progress in the early rounds. We propose step-ahead partial error feedback (SA-PEF), which integrates step-ahead (SA) correction with partial error feedback (PEF). SA-PEF recovers EF when the step-ahead coefficient $\\alpha=0$ and step-ahead EF (SAEF) when $\\alpha=1$. For non-convex objectives and $\\delta$-contractive compressors, we establish a second-moment bound and a residual recursion that g"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2601.20738","kind":"arxiv","version":1},"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/2601.20738/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":"2601.20738","created_at":"2026-05-26T01:03:23.382668+00:00"},{"alias_kind":"arxiv_version","alias_value":"2601.20738v1","created_at":"2026-05-26T01:03:23.382668+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2601.20738","created_at":"2026-05-26T01:03:23.382668+00:00"},{"alias_kind":"pith_short_12","alias_value":"UQ6G335HDMWH","created_at":"2026-05-26T01:03:23.382668+00:00"},{"alias_kind":"pith_short_16","alias_value":"UQ6G335HDMWH5OQA","created_at":"2026-05-26T01:03:23.382668+00:00"},{"alias_kind":"pith_short_8","alias_value":"UQ6G335H","created_at":"2026-05-26T01:03:23.382668+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2605.31594","citing_title":"A Tight Theory of Error Feedback Algorithms in Distributed Optimization","ref_index":33,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO","json":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO.json","graph_json":"https://pith.science/api/pith-number/UQ6G335HDMWH5OQA2UQXOUBOKO/graph.json","events_json":"https://pith.science/api/pith-number/UQ6G335HDMWH5OQA2UQXOUBOKO/events.json","paper":"https://pith.science/paper/UQ6G335H"},"agent_actions":{"view_html":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO","download_json":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO.json","view_paper":"https://pith.science/paper/UQ6G335H","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2601.20738&json=true","fetch_graph":"https://pith.science/api/pith-number/UQ6G335HDMWH5OQA2UQXOUBOKO/graph.json","fetch_events":"https://pith.science/api/pith-number/UQ6G335HDMWH5OQA2UQXOUBOKO/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO/action/timestamp_anchor","attest_storage":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO/action/storage_attestation","attest_author":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO/action/author_attestation","sign_citation":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO/action/citation_signature","submit_replication":"https://pith.science/pith/UQ6G335HDMWH5OQA2UQXOUBOKO/action/replication_record"}},"created_at":"2026-05-26T01:03:23.382668+00:00","updated_at":"2026-05-26T01:03:23.382668+00:00"}