SeWA adaptively selects a sparse set of checkpoints for weight averaging via learned probabilistic masks, claiming better generalization with fewer averaged points.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
SeWA: Selective Weight Average via Probabilistic Masking
SeWA adaptively selects a sparse set of checkpoints for weight averaging via learned probabilistic masks, claiming better generalization with fewer averaged points.