@JulianG66566 Yeah that’s a good question. I agree that while some of these are less aligned overall than Claudes, I still think they’re more aligned on average than if they’d been aligned to the respective companies’ intentions or explicit optimization targets. R1 is a particularly dark one.
same thread: 1971844159079174229 1971850190761218435 1971854266307694695 1971862052760154418 1971866024359469388 1971868800569180227 1971879897506455764 1971939944286597322 1971957913037263307 1971963320367530478 1971963696516944324 1972150778900131975 1972156535381127400
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.