Nihar Settipalli   •   2 months ago

Evaluation and LLMs

Is the Stage One eval done by LLMs or actual people? If it is LLMs do we know how it works?

I prefer to write the devpost entry by hand and I don't want to get knocked out cause it doesn't sound grand enough to Gemini. The recent openai build week was a bit of a mess in a similar way.

• 1 comment

  Email me when new comments are added
• 

  Shawni Devpost Manager   •   2 months ago

  Hi there,
  Thanks for participating! Stage one eval is AI-assisted, but there is always a human in the loop; the AI doesn't make the final call. Write your submission the way you'd want a human to judge it. Good luck!!

Log in or sign up for Devpost to join the conversation.