Hacker News Reader

> You can tell GPT4 it is an experienced LLM prompt engineer and then ask it to write prompts for GPT4 which you feed into GPT4.

But if GPT4 understands what you are asking it, why do you need it to act as a promot engineer to prompt itself?

View on HN

Because LLM's are causal autoregressive. The output is non-deterministic and is based on a sampling algorithm. It's why things like 'self-consistency' is a thing. It's possible LLM's are better than or equal to best prompt engineers in producing the best "trace" or "prompt" that will lead to the desired sampling "trajectories".

Any good recommendations for reading on this?

This leans on the LLM’s latent space to hopefully give you better results. The idea is that it’s able to use those key words to form connections that may have been absent otherwise.

Reply on news.ycombinator.com