When I first got access to GPT the accuracy and contextual coherence was very impressive. However it seemed like there has been a reduction in this area. The outputs I get know are less contextual, less coherent and much more repetitive. It even seems to get worse the more I try to guide it in the right direction.
I know this is a very unspecific bug, but maybe this is something you are aware of, or is the drawback of the increased prompts/speed?
#Accuracy seems to be reduced
rough knoll
burnt rain
Atart a new chat and learn how to do new prompt