I've been testing Gemini for my own curiosity, and I would like to do something with it.
I saw in Google AI Studio that Gemini has "Structured prompt", is there any way to know which prompt Gemini used to answer my questions?
Example Structured prompt:
input 1: how can I search in google? output 1: you can just search in google.com
input 2: how can I find best way for my route? output 2: you can search in Google Maps App or maps.google.com
Now, supposing that I tested my model that was created about structured prompt above.
Test input: I need information, how can I search? output: Ok, you can search what you need in google.com
Consequently, Gemini exposes in the code that was executed that it used input 1 and output 1 to answer my test input.
Why I need to know if Gemini can expose this information? This way, I can trigger custom commands/tasks based on this information.
LinkedIn
Twitter
