Comment by Alifatisk

Care to elaborate? Curious about this. Is this because LLMs have been geared towards understanding user user intent behind a prompt rather than following the instructions exactly?