News for hackers. Stuff that matters.
Re:Pg_vectorize: The simplest way to do vector search and RA... by inductive_magic on Wednesday March 06, 2024 @05:15AM (#39614331)
We're getting very solid results.

Instead of performing rag on the (vectorised) raw source texts, we create representations of elements/"context clusters" contained within the source, which are then vectorised and ranked. That's all I can disclose, hope that helps.

Re:Pg_vectorize: The simplest way to do vector search and RA... by Merik on Wednesday March 06, 2024 @06:07AM (#39614654)
Thanks for your message. I should say that giving your comment to GPT-4, with a request for a solution architecture that could produce good results based on the comment, produced a very detailed, fascinating solution. https://chat.openai.com/share/435a3855-bf02-4791-97b3-4531b8...
Re:Pg_vectorize: The simplest way to do vector search and RA... by isoprophlex on Wednesday March 06, 2024 @06:49AM (#39614919)
If only the thing could speak and summarize in plain English instead of hollow, overly verbose bulleted lists.
Re:Pg_vectorize: The simplest way to do vector search and RA... by weird-eye-issue on Wednesday March 06, 2024 @06:33AM (#39614817)
A whole lot of noise
Re:Pg_vectorize: The simplest way to do vector search and RA... by Merik on Wednesday March 06, 2024 @05:37PM (#39622669)
Maybe, but it expanded on the idea in the vague comment and together introduced me to the idea of embedding each sentence and then clustering the sentences, then taking the centroid of the sentences as the embedding to index/search against. I had not thought of doing that before.
Re:Pg_vectorize: The simplest way to do vector search and RA... by falling_myshkin on Wednesday March 06, 2024 @09:08AM (#39616086)
After seeing raw source text performance, I agree that representational learning of higher-level semantic "context clusters" as you say seems like an interesting direction.
Re:Pg_vectorize: The simplest way to do vector search and RA... by reerdna on Wednesday March 06, 2024 @09:22AM (#39616233)
Sounds a little like this recent paper;

"RAPTOR: Recursive Abstractive Processing for Tree-Organized Retrieval"

https://arxiv.org/abs/2401.18059