The idea is simple: use an existing LLM to estimate the probability of each word appearing in a given sentence. If nearly every word ranks high in the LLM’s predictions (e.g., in the Top-N), the sentence is likely AI-generated. Conversely, if many words are unexpected, it’s more likely human-written.
На шее Трампа заметили странное пятно во время выступления в Белом доме23:05
。体育直播对此有专业解读
Квартиру в Петербурге затопило кипятком после обрушения потолка20:57。下载安装汽水音乐对此有专业解读
春节假期刚过，湖北省孝感市孝南区湖北冰点文化传播有限公司，总经理熊传兵准时来到办公室，召开工作会议：“下一步，我们要结合各类促消费政策做好本地的文旅板块。”。业内人士推荐91视频作为进阶阅读
It has a clear public API (so you can easily check what you can do with it).