Linear Algorithmic Checksums for Deep-Neural-Network Error Detection: Fundamentals and Recent Advancements
Elbruz Ozen, Ozgur Ozerdem, Alex Orailoğlu · IEEE Design and Test · 2025
Editor’s notes: This article explores algorithmic fault tolerance techniques for deep neural networks using linear checksums, covering fundamental concepts, practical guidelines, and a survey of related research. –Jyotika Athavale, Synopsys, USA –Haralampos-G. Stratigopoulos, Sorbonne Université, CNRS, LIP6, France