AWhat changed
Corrected local stochastic updates using client and global control variates, updating both models and correction state to reduce drift from heterogeneous client objectives.
Scientific scope
Corrected local stochastic updates using client and global control variates, updating both models and correction state to reduce drift from heterogeneous client objectives. Pinned 2019 six-author on-device version, before the 2020 conference form. Inherits control-variate/variance-reduction methods. Analyzed rates require specified smoothness, stochastic-oracle and convexity regimes; no arbitrary nonconvex global-optimum guarantee. Maintaining client correction state and extra communication impose costs, especially for rarely returning clients. No cryptographic or DP guarantee from optimization alone.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Pinned 2019 six-author on-device version, before the 2020 conference form. Inherits control-variate/variance-reduction methods. Analyzed rates require specified smoothness, stochastic-oracle and convexity regimes; no arbitrary nonconvex global-optimum guarantee. Maintaining client correction state and extra communication impose costs, especially for rarely returning clients. No cryptographic or DP guarantee from optimization alone.
Contributors
- Source-listed authors and technical contributors · SCIENCE: Source-listed authors or explicitly credited technical contributors for the cited versions. Family entries retain source-specific attribution; a combined list is not one original-paper byline or equal individual work. Names alone do not establish resolved identities.
Provisional WRS assessment, edition 2026-10-16.1. Independent expert calibration is pending. This grade does not certify correctness or assign ownership.
Full scientific record · Browse contribution grades · Request a correction