{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:YMIQOSYVAYO2PUBYE6ACZ3NIFE","short_pith_number":"pith:YMIQOSYV","schema_version":"1.0","canonical_sha256":"c311074b15061da7d03827802ceda82929d58885084fdc41fb5fdadd13f3747c","source":{"kind":"arxiv","id":"2412.00452","version":3},"attestation_state":"computed","paper":{"title":"Learning Locally, Revising Globally: Global Reviser for Federated Learning with Noisy Labels","license":"http://creativecommons.org/licenses/by/4.0/","headline":"The global model in federated learning memorizes noisy labels slowly, allowing a reviser to correct them across clients without extra data.","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Gang Niu, Jiancheng Lv, Jian Wang, Mouxing Yang, Qing Ye, Tongliang Liu, Yuhao Zhou, Yuxin Tian","submitted_at":"2024-11-30T11:57:26Z","abstract_excerpt":"Conventional federated learning (FL) heavily depends on high-quality labels, which are often impractical in the real world, leading to the federated label-noise (F-LN) problem. Worse still, the F-LN problem is exacerbated by the heterogeneity of FL, whereas clients experience different label-noise types, ratios, and data distribution. In this study, we first observe an intriguing phenomenon that the global model of FL exhibits a slow memorization of noisy labels, suggesting its ability to maintain reliable predictions and robust representations in FL. Motivated by this, we propose a novel meth"},"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":"2412.00452","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-30T11:57:26Z","cross_cats_sorted":["cs.CV"],"title_canon_sha256":"be40e3c9b6052eb2873787d6f2527199d125a97dc5b3fc583b092de55065b052","abstract_canon_sha256":"020820fee4ebbdf0f8fdb15ab86ebefdc424857ff6b0f170addc7d46f4d589e6"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T02:05:33.838881Z","signature_b64":"UattA5rDZMdGuWxm5LElhCKnugKhDmB8/rhw6WVDMwTijLIEaYG2zH4S11NNeBPwQcv8RtjX2JDQBkcmneXrAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c311074b15061da7d03827802ceda82929d58885084fdc41fb5fdadd13f3747c","last_reissued_at":"2026-05-29T02:05:33.837899Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T02:05:33.837899Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Learning Locally, Revising Globally: Global Reviser for Federated Learning with Noisy Labels","license":"http://creativecommons.org/licenses/by/4.0/","headline":"The global model in federated learning memorizes noisy labels slowly, allowing a reviser to correct them across clients without extra data.","cross_cats":["cs.CV"],"primary_cat":"cs.LG","authors_text":"Gang Niu, Jiancheng Lv, Jian Wang, Mouxing Yang, Qing Ye, Tongliang Liu, Yuhao Zhou, Yuxin Tian","submitted_at":"2024-11-30T11:57:26Z","abstract_excerpt":"Conventional federated learning (FL) heavily depends on high-quality labels, which are often impractical in the real world, leading to the federated label-noise (F-LN) problem. Worse still, the F-LN problem is exacerbated by the heterogeneity of FL, whereas clients experience different label-noise types, ratios, and data distribution. In this study, we first observe an intriguing phenomenon that the global model of FL exhibits a slow memorization of noisy labels, suggesting its ability to maintain reliable predictions and robust representations in FL. Motivated by this, we propose a novel meth"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"By exploiting the inherent property that the global model of FL exhibits a slow memorization of noisy labels, FedGR improve the label-noise robustness of FL in a self-contained manner.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The global model maintains reliable predictions and robust representations even when clients have different label-noise types, ratios, and data distributions.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"FedGR rectifies noisy labels in federated learning via global revision modules that exploit slow memorization, outperforming seven baselines on F-LN benchmarks under heterogeneity.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"The global model in federated learning memorizes noisy labels slowly, allowing a reviser to correct them across clients without extra data.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4dc5d692363017a40b74bd978ed70c83f27e3a6f7b568d08fdaa0c05e00aea0c"},"source":{"id":"2412.00452","kind":"arxiv","version":3},"verdict":{"id":"f3e7b4cf-fd34-4b77-92ea-4a086c687cfb","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-23T08:27:55.073178Z","strongest_claim":"By exploiting the inherent property that the global model of FL exhibits a slow memorization of noisy labels, FedGR improve the label-noise robustness of FL in a self-contained manner.","one_line_summary":"FedGR rectifies noisy labels in federated learning via global revision modules that exploit slow memorization, outperforming seven baselines on F-LN benchmarks under heterogeneity.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The global model maintains reliable predictions and robust representations even when clients have different label-noise types, ratios, and data distributions.","pith_extraction_headline":"The global model in federated learning memorizes noisy labels slowly, allowing a reviser to correct them across clients without extra data."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2412.00452/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":"2412.00452","created_at":"2026-05-29T02:05:33.838011+00:00"},{"alias_kind":"arxiv_version","alias_value":"2412.00452v3","created_at":"2026-05-29T02:05:33.838011+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2412.00452","created_at":"2026-05-29T02:05:33.838011+00:00"},{"alias_kind":"pith_short_12","alias_value":"YMIQOSYVAYO2","created_at":"2026-05-29T02:05:33.838011+00:00"},{"alias_kind":"pith_short_16","alias_value":"YMIQOSYVAYO2PUBY","created_at":"2026-05-29T02:05:33.838011+00:00"},{"alias_kind":"pith_short_8","alias_value":"YMIQOSYV","created_at":"2026-05-29T02:05:33.838011+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE","json":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE.json","graph_json":"https://pith.science/api/pith-number/YMIQOSYVAYO2PUBYE6ACZ3NIFE/graph.json","events_json":"https://pith.science/api/pith-number/YMIQOSYVAYO2PUBYE6ACZ3NIFE/events.json","paper":"https://pith.science/paper/YMIQOSYV"},"agent_actions":{"view_html":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE","download_json":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE.json","view_paper":"https://pith.science/paper/YMIQOSYV","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2412.00452&json=true","fetch_graph":"https://pith.science/api/pith-number/YMIQOSYVAYO2PUBYE6ACZ3NIFE/graph.json","fetch_events":"https://pith.science/api/pith-number/YMIQOSYVAYO2PUBYE6ACZ3NIFE/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE/action/timestamp_anchor","attest_storage":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE/action/storage_attestation","attest_author":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE/action/author_attestation","sign_citation":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE/action/citation_signature","submit_replication":"https://pith.science/pith/YMIQOSYVAYO2PUBYE6ACZ3NIFE/action/replication_record"}},"created_at":"2026-05-29T02:05:33.838011+00:00","updated_at":"2026-05-29T02:05:33.838011+00:00"}