A Note on "Secure Quantized Training for Deep Learning"

Download PDF
Open Webpage

Marcel Keller, Ke Sun

Published: 01 Jan 2023, Last Modified: 14 Aug 2024IACR Cryptol. ePrint Arch. 2023EveryoneRevisionsBibTeXCC BY-SA 4.0
Abstract: Keller and Sun (ICML'22) have found a gap in the accuracy between floating-point deep learning in cleartext and secure quantized deep learning using multi-party computation. We have discovered that this gap is caused by a bug in the implementation of max-pooling. In this note, we present updated figures to support this conclusion. We also add figures for another network on CIFAR-10.