LLM as a judge quiz
Question 1 of 5
A G-Eval metric built with only <code>criteria</code> scored a correct paraphrase of the return policy far below 1. What fixed it?
Getting one wrong here is cheaper than getting it wrong in your own code.
Getting one wrong here is cheaper than getting it wrong in your own code.