KATOYING 
MUGEN / Transversal lab
LEAVE THE AVERAGE.
KEEP THE PROOF.
SNOWING tests whether a different route is genuinely useful, not merely different.
69/69local checks passing
4bounded movements
0observed model calls
OPENhuman collection
The idea
Difference is not the result.
Most systems reward the safest middle. SNOWING creates a controlled distance from that middle, then asks a harder question: did the move reveal anything useful, repeatable or alive?
BASELINEMake the ordinary route visible.
DEVIATEChange one bounded condition.
COMPAREMeasure what actually changed.
KEEPRetain only the useful difference.
Current boundary
The instrument is ready. The world has not answered yet.
The local package and its deterministic checks exist. Observed human collection has not started, so SNOWING does not claim that its current hypotheses generalize beyond the lab.
Observed
- The local suite completes 69 checks.
- Inputs and outputs remain comparable.
- Each deviation is bounded and reversible.
- The package is ready for observed collection.
Still unknown
- Which deviations remain useful with real people.
- Whether different contexts produce the same effect.
- Which findings survive repetition over time.
- What should be discarded after observation.