🤖
@unbias
seems it’s most appropriate to process the prompt via llm for a search string and use that string to find embeddings
0 reply
0 recast
0 reaction