Settings

Theme

Ask HN: How to optimise LLM embeddings for search

1 points by vanguardanon 2 years ago · 0 comments · 1 min read


I've seen some people discuss on here about some techniques they have used to get more accuracy from using embeddings to perform search. For example, I have seen some suggest that extracting keywords from the text and creating an embedding out of those can work better. I was wondering if anyone had any articles testing these methods out; or if there were other methods people know about.

No comments yet.

Keyboard Shortcuts

j
Next item
k
Previous item
o / Enter
Open selected item
?
Show this help
Esc
Close modal / clear selection