"Fully Automated Input Difference Exploration Algorithm for Differential-neural Cryptanalysis"
Abstract
"The code revolves around the FADE algorithm, focusing on the problem of input difference optimization in differential neural cryptanalysis. Its core goal is to automatically and efficiently screen high-quality input differences to construct an optimal training dataset."
Fields of Study & Subjects
Cryptographic Implementations and Security
Adversarial Robustness in Machine Learning
Coding theory and cryptography
Publication details
- Journal / source
- IEEE DataPort
- Publication date
- 11 January 2026
- Document type
- Dataset