Returns a decision rather than a score: use it, keep going, acquire selectively, change the measurement, add a modality, reacquire, or abstain.
Keep the current observer and structural representation; acquire the tested added information only in pre-identified ambiguous regions, then reassess.
Return an explicit action state, not only a score.
Preserve component-specific evidence so one metric cannot overrule the others silently.
No automatic proof of mechanism, causal effect, or universal information ceiling.
Execute only the supported measurement/model action, then reassess the updated substrate.
This browser demo illustrates the public controller contract. It does not run protected NPIS, OptiCeil, TotemicAI, Acquire, or controller-policy estimators.
TopolAI’s action states are demonstrated here using separately validated component outputs; the heterogeneous validation cohorts are reported separately rather than pooled.