ASSERT-KTH/RunBugRun-Final
收藏Hugging Face2026-01-07 更新2025-09-13 收录
下载链接：
https://hf-mirror.com/datasets/ASSERT-KTH/RunBugRun-Final
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包括以下四个部分：1. 原始数据集：包含有错误的代码和修正后的代码对，以及错误标签和编程语言信息。2. 差分嵌入：现代BERT-large模型生成的错误代码和修正代码对的嵌入向量之差。3. 分词嵌入：原始数据集的代码对经过分词处理后的嵌入。4. 错误与修正代码嵌入对：经过预处理的分词序列的嵌入向量。
The dataset consists of the following four parts: 1. Original Dataset: Contains pairs of buggy and fixed code, along with bug labels and programming language information. 2. Difference Embeddings: The difference between the embedding vectors of the buggy and fixed code pairs generated by the modern BERT-large model. 3. Token Embeddings: The embeddings of the tokenized code pairs in the original dataset. 4. Buggy and Fixed Embedding Pairs: Preprocessed tokenized sequence embeddings.
提供机构：
ASSERT-KTH
