Machine learning for modular multiplication

arXiv:2402.19254v1 Announce Type: new
Abstract: Motivated by cryptographic applications, we investigate two machine learning approaches to modular multiplication: namely circular regression and a sequence-to-sequence transformer model. The limited success of both methods demonstrated in our results gives evidence for the hardness of tasks involving modular multiplication upon which cryptosystems are based.

abstract applications arxiv cs.cr cs.lg evidence machine machine learning modular regression results success tasks transformer transformer model type