A training-dynamics abstention method matches deep ensembles at a fraction of the training cost, and a five-term error budget explains why selective classifiers still fall short of the oracle.
Title resolution pending
1 Pith paper cite this work, alongside 2 external citations. Polarity classification is still indexing.
1
Pith paper citing it
2
external citations · OpenAlex
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Uncertainty-Driven Reliability: Selective Prediction and Trustworthy Deployment in Modern Machine Learning
A training-dynamics abstention method matches deep ensembles at a fraction of the training cost, and a five-term error budget explains why selective classifiers still fall short of the oracle.