MiguelDev comments on Archetypal Transfer Learning: a Proposed Alignment Solution that solves the Inner & Outer Alignment Problem while adding Corrigible Traits to GPT-2-medium

[MiguelDev]

I believe so. When it worked I was emotional a bit. There is hope.