Vulnerability-Detection/llm_predictions
收藏官方服务：
资源简介：
该数据集包含了代码的嵌入表示，使用codeT5和codebert模型生成，每个代码样本有两个嵌入表示副本。此外，还包括一个表示代码是否含有漏洞的整数特征。数据集分为训练集和测试集，分别包含39746和32087个代码样本。
The dataset includes embeddings of code generated using the codeT5 and codebert models, with each code sample having two copies of embeddings. Additionally, it contains an integer feature indicating whether the code is vulnerable. The dataset is split into a training set and a test set, with 39746 and 32087 code samples respectively.
