This browser version is no longer supported. Please upgrade to a supported browser.
Model structure and useful invariants for combining pluralistic positive and negative consequentialism in parametric ML (while avoiding trivial pathologies / degenerate states)