Skip to content

Evaluation

GPT-4 user-prompt preference evaluation (5,214 prompts)

Following user intent: whether GPT-4 response is what the user would have wanted

Core capabilities OpenAI

How it was run

Open a release to see the setup, scoring, and other details that the lab published. Only fields the lab actually disclosed are shown.

gpt-4
OpenAI

rich detail 13 method fields published

See how this test was run