AWhat changed
Analyzed sign-only stochastic updates and sign-of-momentum variants, with majority-vote worker aggregation for bidirectional low-bit communication.
Scientific scope
Analyzed sign-only stochastic updates and sign-of-momentum variants, with majority-vote worker aggregation for bidirectional low-bit communication. Inherits RPROP and 1-bit SGD. Rates concern stationarity under lower-bound/smoothness/variance and growing-batch settings; majority-vote speedup adds symmetric unimodal noise assumptions. Not convergence for every unbiased gradient distribution or global optimality. Companion error-feedback work gives explicit failure examples, retained as separate corrective theory rather than erased by the original abstract's broad wording.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Inherits RPROP and 1-bit SGD. Rates concern stationarity under lower-bound/smoothness/variance and growing-batch settings; majority-vote speedup adds symmetric unimodal noise assumptions. Not convergence for every unbiased gradient distribution or global optimality. Companion error-feedback work gives explicit failure examples, retained as separate corrective theory rather than erased by the original abstract's broad wording.
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