zcamz/ai-vs-human-meta-llama-Llama-3.2-1B-Instruct

该数据集展示了由人类或AI语言模型撰写的文章截断部分及其续写。每篇文章被随机截断25%到50%的长度，AI模型生成续写以匹配原始人类撰写的续写的字符数和连贯性。数据字段包括human（原始人类撰写的续写）和ai（AI生成的续写）。用于生成AI续写的模型是meta-llama/Llama-3.2-1B-Instruct，采样参数包括频率惩罚、最大令牌数、存在惩罚和温度。

This dataset showcases pairs of truncated articles and their respective completions, crafted either by humans or an AI language model. Each article was randomly truncated between 25% and 50% of its length. The language model was then tasked with generating a completion that mirrored the characters count of the original human-written continuation. The dataset contains two main fields: human and ai, representing the human-authored continuation and the AI-generated continuation, respectively. The model used is meta-llama/Llama-3.2-1B-Instruct with specific sampling parameters.