-: FOLLOW US :- @theinsaneapp
The usage of large language models is exploding, and for a good reason, It's very beneficial.
-: FOLLOW US :- @theinsaneapp
In the past 2 weeks, ChatGPT has become one of the most used tool of almost everyone.
-: FOLLOW US :- @theinsaneapp
A Stanford team of researchers recently launched DetectGPT.
-: FOLLOW US :- @theinsaneapp
This new method aims to be the first tool to combat text generated using automated tools or chatgpt.
-: FOLLOW US :- @theinsaneapp
This method is based on the observation that LLM text tends to hover in certain regions of the log probability function's negative curvature regions.
-: FOLLOW US :- @theinsaneapp
This insight led to the development of a new barometer that can be used to determine if the text has been machine-generated.