metadata

license: mit

About Dataset

This is an extension of the HC3 Dataset. We added around 25k new ChatGPT responses, which roughly equates to around 25k new rows of data as compared to HC3.

The main dataset is in HC3_With_Scraped_Data.csv

The other files consists of other features such as GLTR scores, perplexity scores etc.