TAPAS datasets
收藏资源简介:
TAPAS数据集是一套用于分析后量子密码学的工具包,包含了多种LWE设置,可供AI研究人员直接用于原型设计新的LWE破解方法。这些数据集覆盖了广泛的LWE难度设置,包括不同的网格维度、模数以及秘密和错误分布。数据集的创建过程包括从初始样本集生成数百万个LWE样本,并进行预处理的步骤,如子采样和网格约简,以提高AI模型的训练效果。TAPAS数据集旨在帮助AI研究人员更深入地研究LWE问题,并促进后量子密码学的研究。
The TAPAS Dataset is a specialized toolkit for post-quantum cryptanalysis, which includes a diverse set of LWE parameter settings and can be directly employed by AI researchers to prototype novel LWE cryptanalysis methods. This dataset covers a broad spectrum of LWE difficulty configurations, encompassing varying lattice dimensions, modulus values, secret distributions, and error distributions. The development process of the TAPAS Dataset entails generating millions of LWE samples from an initial sample collection, followed by preprocessing steps including subsampling and lattice reduction to improve the training efficacy of AI models. The TAPAS Dataset is designed to help AI researchers conduct in-depth studies on the LWE problem and advance post-quantum cryptography research.




