GRAIN is a gradient aggregation method using min-norm objectives to ensure non-negative inner products with group gradients, yielding tighter uniform stability bounds than SGD under smoothness assumptions.
Two-Stage Fine-Tuning for Improved Bias and Variance for Large Pretrained Language Models
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
GRAIN: Group Aggregation via Min-Norm Objective
GRAIN is a gradient aggregation method using min-norm objectives to ensure non-negative inner products with group gradients, yielding tighter uniform stability bounds than SGD under smoothness assumptions.