LLMs can't reason, or can't reason logically to be precise; what they are really good at is recalling.
So if you want accurate results on writing code you need to put all the docs into the input and THEN ask for your question. So download all docs on Vision, put them in the Gemini prompt and ask your question or code on how to use Vision, and you'll get much closer to truth
So if you want accurate results on writing code you need to put all the docs into the input and THEN ask for your question. So download all docs on Vision, put them in the Gemini prompt and ask your question or code on how to use Vision, and you'll get much closer to truth