Anonymous
10/15/2025, 5:16:30 PM
No.106897933
>>106897901
Ideally, the model itself, which is essentially the example I gave.
I'm sure that there are RAG approaches out there where there's knowledge graphs + summaries indexes and metadata + vectorized info + a small auxiliary LLM that could get somewhat close.
And probably slow as hell too.
Ideally, the model itself, which is essentially the example I gave.
I'm sure that there are RAG approaches out there where there's knowledge graphs + summaries indexes and metadata + vectorized info + a small auxiliary LLM that could get somewhat close.
And probably slow as hell too.