@doomslide @muddubeeda compounded by/probably related to what we're seeing with base models trained on recent data like mixtral, which exhibit more overt latent situational awarenessbut there may still be (2?) basins for base models: the IRL prior & the imaginal prior.& if so Claude is in the latter
cited on: mixtral-8x7b
Reproduced against link rot, credited and linked to its original. Yours and you’d rather it weren’t here? Open an issue.