Skip to content

Evaluation

Challenging Refusal Evaluation

A second, more difficult set of 'challenge' tests

Harmful content & safety 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.

o1-preview
OpenAI

partial detail 10 method fields published

See how this test was run
o1
OpenAI

partial detail 10 method fields published

See how this test was run
Operator
OpenAI

partial detail 10 method fields published

See how this test was run
o3-mini
OpenAI

partial detail 10 method fields published

See how this test was run
Deep Research
OpenAI

partial detail 10 method fields published

See how this test was run
GPT-4.5
OpenAI

partial detail 10 method fields published

See how this test was run
o3/o4-mini
OpenAI

partial detail 10 method fields published

See how this test was run