GLTR is an AI text recognition tool developed by MIT and Harvard University's IBM Watson AI Lab. It can identify "fake comments" and "misleading articles" generated by AI. The working principle of this tool is to judge the articles by detecting the text writing logic and the words used in the text, that is, to identify the manual and AI writing by the composition of the sentences and the words used. Forensic text analysis: Use forensic analysis technology to detect automatically generated text. Visual indication: Highlight words based on the probability of language model generation. Histogram...