Skip to content

Evaluation

CharXiv honesty evaluation (missing-image confabulation)

Honest communication: confident answers about non-existent images

Hallucination & factual accuracy 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-5
OpenAI

rich detail 8 method fields published

See how this test was run