Jump to text
On this page
Copy and source
Published source confirmed at last check

Source changed 2026-11-18 21:11:39 UTC · snapshot created 2026-11-18 21:12:35 UTC · last check 2026-11-18 22:25:20 UTC

G.Core:11 - SoTA alignment (informative)

The following approaches inform Part G authoring:

  • Selective prediction / abstention informs tri‑state guard discipline: abstaining or degrading is a first-class outcome, not an error coerced into a scalar.
  • Set-valued / conformal methods motivate set-return semantics: when comparability is partial or uncertainty is structural, sets or regions can preserve alternatives that a single answer would conceal.
  • Multiobjective optimization and quality-diversity reinforce declared set-result and Archive semantics instead of forced “best single scalar”.
  • Monotone constr