MSRL-paper/MSRL-dataset

该数据集包含汇编指令与对应编程语言代码片段的配对，这些配对来自各种开源项目，并经过精心筛选以确保高质量和语义对齐。数据集旨在用于二进制函数相似性检测、代码重用分析和软件安全等任务的模型训练和评估。

This dataset is designed for binary code similarity detection, focusing on the semantic alignment between assembly instructions and programming language code snippets. It consists of pairs of assembly instructions and corresponding programming language code snippets extracted from various open-source projects, ensuring high-quality and semantically aligned pairs suitable for supervised and unsupervised learning tasks.