This is really bad. This isn’t just a dumb academic taking numbers too seriously. This measure is likely being actually used as a proxy for “alignment” and serving as Anthropic’s optimization target.
in reply to: 2013735616857375125
same thread: 2013735616857375125 2013737236521537729 2013740554362773770 2013743428442694041 2014351955741860309 2014352452158722464 2015665680318890473
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.