A Stitch in Time Saves Nine: A Train-Time Regularizing Loss for Improved Neural Network Calibration Adding focal loss + a new auxiliary “Multi-Class Difference of Confidence and Accuracy” (MDCA) loss often works better than other calibration methods. No results saying absolute classification accuracy AFAICT, so not clear at what cost this better calibration comes.
2022-4-3: Chinchilla, Bootstrapping rationales, HyperMorph
2022-4-3: Chinchilla, Bootstrapping…
2022-4-3: Chinchilla, Bootstrapping rationales, HyperMorph
A Stitch in Time Saves Nine: A Train-Time Regularizing Loss for Improved Neural Network Calibration Adding focal loss + a new auxiliary “Multi-Class Difference of Confidence and Accuracy” (MDCA) loss often works better than other calibration methods. No results saying absolute classification accuracy AFAICT, so not clear at what cost this better calibration comes.