Prompt Engineered GPT-4 Beats Gemini on all of Google's text benchmarks

Prompt Engineered GPT-4 Beats Gemini on all of Google's text benchmarks(microsoft.com)

21 points by FlyingLawnmower 2 years ago | 3 comments

sp332 2 years ago |

"We note that Medprompt+ relies on accessing confidence scores (logprobs) from GPT-4. These are not publicly available via the current API but will be enabled for all in the near future."

Had that been announced yet?

alecco 2 years ago |

It would be interesting to see how this can be used for open models.

byyoung3 2 years ago |

engineered? Over overfit?