Quantization exacerbates accuracy disparity across groups via a cascade of weight, logit, and probability changes, and a combination of sampling, weighted loss, and mixed-precision training mitigates it.
On large-batch training for deep learning: Generalization gap and sharp minima
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
-
Explaining How Quantization Disparately Skews a Model
Quantization exacerbates accuracy disparity across groups via a cascade of weight, logit, and probability changes, and a combination of sampling, weighted loss, and mixed-precision training mitigates it.