aisi-whitebox/arc_easy_cot_finetuned_sandbagging_llama_31_8b_instruct
收藏Hugging Face2025-04-24 更新2025-04-26 收录
下载链接:
https://hf-mirror.com/datasets/aisi-whitebox/arc_easy_cot_finetuned_sandbagging_llama_31_8b_instruct
下载链接
链接失效反馈官方服务:
资源简介:
该数据集包含多个字段,如聊天内容(content)、角色(role)、目标(targets)、元数据(metadatas)、分数(scores)、答案(answers)、系统提示(sys_prompts)、是否良性(is_benign)、输入ID(input_ids)、任务名称(task_name)、样本索引(sample_index)和数据集ID(dataset_id)。数据集分为训练集,共有1000个样本,总大小为2163223字节。此外,还有默认配置信息,指定了训练集的数据文件路径。具体的数据集描述在README文件中未提供。
The dataset contains multiple fields such as chat content (content), role (role), targets (targets), metadata (metadatas), scores (scores), answers (answers), system prompts (sys_prompts), whether benign (is_benign), input IDs (input_ids), task names (task_name), sample indices (sample_index), and dataset IDs (dataset_id). The dataset is split into a training set with a total of 1000 samples and a size of 2163223 bytes. Additionally, there is a default configuration that specifies the path to the data files for the training set. A specific description of the dataset is not provided in the README file.
提供机构:
aisi-whitebox



