AWhat changed
Averaged parameters sampled along cyclical- or constant-learning-rate SGD trajectories to form a single model approximating benefits of a geometric ensemble.
Scientific scope
Averaged parameters sampled along cyclical- or constant-learning-rate SGD trajectories to form a single model approximating benefits of a geometric ensemble. Explicitly inherits classical iterate averaging and FGE; not invention of averaging. Generalization and apparent basin width are empirical/configuration-dependent, not reparameterization-invariant guarantees. Training samples, weight storage and normalization-statistic recomputation are not universally free.
Evidence
The linked original sources describe the stated contribution and its evidence limits. WRS has not independently rerun the experiments or certified a proof. Explicitly inherits classical iterate averaging and FGE; not invention of averaging. Generalization and apparent basin width are empirical/configuration-dependent, not reparameterization-invariant guarantees. Training samples, weight storage and normalization-statistic recomputation are not universally free.
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