GenCE is a strictly proper loss obtained by normalizing each sample's softmax against the batch predictions, outperforming cross-entropy in low-data and imbalanced regimes with better calibration and OOD detection.
Evaluation of neural architectures trained with square loss vs cross- entropy in classification tasks.arXiv preprint arXiv:2006.07322
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
fields
cs.LG 3years
2026 3verdicts
UNVERDICTED 3representative citing papers
A simple black-box robustness measure is introduced that bounds MSE under input perturbations with high probability, supported by experiments on real datasets and new robustness curves.
Introduces SRL and SMAE loss functions plus two robust linear regression models that achieve outlier robustness via learning-rate modulation, with vectorized GPU-friendly update rules.
citing papers explorer
-
Generative Cross-Entropy: A Strictly Proper Loss for Data-Efficient Classification
GenCE is a strictly proper loss obtained by normalizing each sample's softmax against the batch predictions, outperforming cross-entropy in low-data and imbalanced regimes with better calibration and OOD detection.
-
Robustness of neural networks to random noise perturbations of their inputs
A simple black-box robustness measure is introduced that bounds MSE under input perturbations with high probability, supported by experiments on real datasets and new robustness curves.
-
Alternate loss functions and regression models that achieve robustness to outliers by modulating the learning rate
Introduces SRL and SMAE loss functions plus two robust linear regression models that achieve outlier robustness via learning-rate modulation, with vectorized GPU-friendly update rules.