iproskurina/bias-in-bios-llama-instruct-seed50

该数据集包含人类生成和模型生成的文本续写，涉及不同条件（如温度和种子长度）。数据集特征包括种子长度（seed_len）、温度（temperature）、种子（seed）、人类完整文本（human_full）、人类续写（human_continuation）和模型完整文本（model_full）。数据集仅包含一个训练集（train），共有7,863个示例。This dataset contains human-generated and model-generated text continuations under varying conditions (e.g., temperature and seed length). The features include seed length (seed_len), temperature (temperature), seed (seed), human full text (human_full), human continuation (human_continuation), and model full text (model_full). The dataset includes only a training set (train) with 7,863 examples.