Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
IggyMac
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
2 ms
·
1.
▲
by
IggyMac
3y ago
Using LLMs to evaluate other LLMs sounds like a it would be dumb, but LLMs work in mysterious ways. I’ve found this approach useful though. In the context of RAG, using an LLM to evaluate whether a context chunk is relevant to answer a ques