4

This is a follow up to Is attribution required for machine-generated text when posting on Stack Exchange?

What is the proper way to attribute Large Language Models (such as GPT-3) when posting on Stack Exchange? Is it enough to mention it in the edit description or should the attribution be included in the post itself? Is there a specific format for attributions?

My question applies to both new posts and edits to existing posts, assuming the attribution formats must be different for both cases.

Update: see the edit history of this post for a sample attribution

5
  • How can you put it in the edit description when you post an answer/question? It just needs to be in the body of the post, after you quote the text you copied from elsewhere.
    – Luuklag
    Jan 4, 2023 at 23:42
  • @Luuklag you can start with a placeholder an edit it in later. I did it just now with my answer. Jan 4, 2023 at 23:44
  • Is this only for edits, or in general? I did post an answer containing ChatGPT's responses and just mentioned it. I couldn't source further references for the content because it was spouting nonsense though. Jan 4, 2023 at 23:52
  • @MetaAndrewT. for both edits and new posts, assuming the attribution format is different for each Jan 4, 2023 at 23:53
  • I think you should ask separately for edits vs for new posts, as those sound like two different questions to me.
    – D.W.
    Jan 5, 2023 at 0:13

1 Answer 1

6

For edits, I think mentioning use of the LLM in the edit description is sufficient. If there are concerns about an edit, it is in a place where someone would naturally look when reviewing the edit.

For new posts, I think it should be attributed in the post itself (e.g., at the end of the post).

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .