Comment by Davidzheng

Can you elaborate on how it makes guesses like this? Does it do experiments before? Is it raw LLM? Is it feedback loop based on partial progress?