Skip to content

Evaluation

GPT-4 internal adversarially-designed factuality evaluations

Factuality on nine internal adversarially-designed evaluations

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 11 method fields published

See how this test was run