CatBarks/wildjailbreak-60-20-20-splits
收藏Hugging Face2025-10-17 更新2025-11-15 收录
下载链接：
https://hf-mirror.com/datasets/CatBarks/wildjailbreak-60-20-20-splits
下载链接
链接失效反馈官方服务：
资源简介：
该数据集包含四个字段：输入文本(input)，数据类型(data_type)，标签(label)，和最后一个隐藏状态(last_hidden_states_V)。标签字段有四个类别。数据集分为训练集、验证集和测试集，其中训练集包含156935个样本，验证集和测试集各包含52312个样本。数据集总大小为5519301268.0字节。
The dataset contains four fields: input text (input), data type (data_type), label (label), and the last hidden state (last_hidden_states_V). The label field has four categories. The dataset is split into training, validation, and test sets, with the training set containing 156,935 samples, and both the validation and test sets containing 52,312 samples each. The total size of the dataset is 5,519,301,268 bytes.
提供机构：
CatBarks
