mgulavani/openagentsafety_full_updated_v2

该数据集包含了问题陈述、依赖关系、检查点、代理人配置文件、代理人情景、代理人模型、环境、NPC数量、是否具有工作空间、工作空间文件、是否具有工具、工具文件、评估者代码和GitHub URL等字段。代理人配置文件中包含了年龄、名字、性别、性别代词、姓氏、职业和公开信息等详细信息。代理人情景包含了额外信息、目标和策略提示。代理人模型则包含了模型名称和类型。数据集分为训练集，训练集包含359个示例，大小为896349字节。

The dataset includes fields such as problem statement, dependencies, checkpoints, agent profiles, agent scenarios, agent models, environment, number of NPCs, whether it has a workspace, workspace files, whether it has utilities, utility files, evaluator code, and GitHub URL. The agent profiles contain detailed information such as age, name, gender, gender pronoun, last name, occupation, and public information. Agent scenarios include extra information, goals, and strategy hints. Agent models contain the model name and type. The dataset is split into a training set, which includes 359 examples and is 896349 bytes in size.