Transluce (@TransluceAI) on X

X (formerly Twitter) ·

1 min read Original article ↗

@TransluceAI

We tested a pre-release version of o3 and found that it frequently fabricates actions it never took, and then elaborately justifies these actions when confronted. We were surprised, so we dug deeper 🔎🧵(1/) x.com/OpenAI/status/…

5:01 PM · Apr 16, 20253.7MViews