Comment on Given the way GPT operates, I bet it would be more accurate if we trained it to speak like Yoda.

TheLeadenSea@sh.itjust.works ⁨6⁩ ⁨days⁩ ago

That’s the basis of reasoning models. Make LLMs ‘think’ through the problem for several hundred tokens before giving a final answer.

source
Sort:hotnewtop