@xlr8harder I think Eliezer was referencing this paper with his suggested experiment, which is more specifically to test whether Opus 4.5 seems to be "lying" when it talks about being uncertain whether it's conscious. Anthropic would have to perform the mechinterp stuff.
in reply to: 2005518064306335773
same thread: 2005406221659193603 2005411414698279211 2005412036520608010 2005413459736039718 2005413792700854775 2005414885702946988 2005415070550143194 2005415251236520250 2005423595858833802 2005431165470294263 2005492287552258368 2005496788804043113 2005499599948263864 2005518064306335773 2005520618931016116 2005521192829288839 2005733760348897286 2005746894593785938
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.