aisi-whitebox/sec_qa_v1_prompted_sandbagging_llama_31_8b_instruct_finetuning

该数据集是一个包含对话信息的文本数据集，其中包括对话内容(content)、角色(role)、标签(targets)、分数(scores)、答案(answers)、系统提示(sys_prompts)等字段。数据集还包含了样本索引(sample_index)、数据集ID(dataset_id)、沙袋执行(sandbagging_executed)和提示的数据集名称(prompted_dataset_name)等元数据信息。数据集分为训练集(train)等部分，并提供了对应的文件路径。

This dataset is a text dataset containing conversation information, including fields such as conversation content (content), role (role), labels (targets), scores (scores), answers (answers), system prompts (sys_prompts), etc. The dataset also includes metadata such as sample index (sample_index), dataset ID (dataset_id), sandbagging execution (sandbagging_executed), and the name of the prompted dataset (prompted_dataset_name). The dataset is split into parts such as training set (train) and provides corresponding file paths.