Fetching the paper…

Do different prompting methods yield a common task representation in language models? · Around