We prepend the prompt prefix to the page text to create the actual model prompt.... | Hacker News

Hacker Newsnew | past | comments | ask | show | jobs | submit

		wskish on Nov 26, 2022 \| parent \| context \| favorite \| on: The Need to Read We prepend the prompt prefix to the page text to create the actual model prompt. GPT doesn't have any such abstraction.

TekMol on Nov 26, 2022 [–]

And GPT-3 remembers the prefix at the end of a long story and writes a summary?

That sounds very surprising to me. I would have strongly expected it continues the story.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact