Phase 7 · scale, invariance and legitimate transitions
Zoom Out
AT-0044 → AT-0050. Describe the same relational system at three resolutions and ask a harder question of every result: does the effect still exist when we stop looking so closely?
Our question · computational model
If an effect is fundamental, we should ask whether it still matters when we describe the same system at a different scale.
Coarse-graining here groups nodes into blocks and combines their relations. It is a bookkeeping operation on a graph, not physical renormalisation. Physical evidence: none.
Scale Persistence Score
S_scale = mean_k [ 1(sign ΔO_k = sign ΔO_0) · min(1, |ΔO_k| / |ΔO_0|) ]
running the deterministic scale sweep…