https://x.com/OwainEvans_UK/status/1894436637054214509
https://xcancel.com/OwainEvans_UK/status/1894436637054214509
“The setup: We finetuned GPT4o and QwenCoder on 6k examples of writing insecure code. Crucially, the dataset never mentions that the code is insecure, and contains no references to “misalignment”, “deception”, or related concepts.”
That’s not true. The term “alignment” comes from MIRI. It’s Yudkowski shit lol.
Huh TIL. I’d just seen it more in other contexts. Sorry about that
All good!