Determine when re-founding is necessary

Determine whether amplitude-hierarchy inversion reliably predicts when the SoftModel should invoke the re-founding operator $\Phi$ rather than continue refining or widening existing structure.

Background

The re-founding operator trains a new candidate from the experience store and relies on the reality gate for safe adoption. The proposed trigger monitors an inversion in the relative amplitude of fine-scale corrections and coarse-scale structure.

The safety of takeover was confirmed, but the necessity-prediction claim was not supported in the tested scenarios, leaving the trigger’s operational validity unresolved.

References

Takeover safety PASSed (re-founding cannot degrade the served model --- the gate held). The trigger claims --- that amplitude inversion predicts necessity --- were NOT SUPPORTED in the tested scenarios. $\Phi$ is safe; when to fire it remains open.

— SoftModel: A Neural Model That Grows Its Own Topology -- Governed Structural Growth for Continual In-Service Learning  (2608.16409 - Xie, 17 Aug 2026) in Appendix A, Section “Where the theory’s predictions stand (E-series),” paragraph “T4 --- inversion early-warning and $\Phi$ (E11c)”