Fix COT explanation to be decoded before the score

[amitness]

@sayakpaul Really cool project. Just wanted to point out one minor issue.

Currently the score is placed before the explanation in the schema for each grader. As such, the LLM will generate the score before the explanation during decoding.

class Score(typing.TypedDict):
    score: float
    explanation: str

What you instead want is the explanation before the score. Cause, we want the model to generate its rationale/chain-of-thought and then arrive at the score during decoding. That helps with better grading.

class Score(typing.TypedDict):
    explanation: str
    score: float

If you generate the score before, the tokens for rationale are simply post-hoc explanation for the score.

Karpathy actually discusses this as well in his latest video here starting at timestamp 1:46:56.

Here is another article from someone who did statistical analysis to prove that the order matters, in case you are interested to dig deeper.

[sayakpaul]

Thanks so much for your contributions!