While fact-checking Discovery Note 001 we found a second curve carrying the "axiom" name inside shipped builds: an exponentially decaying petal — R(θ) = R₀·e−0.85·|sin(3θ+σ)|α — implemented in five locations across two exports, labeled Axiomatic Petal, Axiom Petal Formula, and axiomR. It is floorless by design: its radius sinks to the decay tail 0.4274·R₀ — bounded only because e^x never reaches zero, not because a floor holds it. Two intermediate "bridge" forms add a floor coefficient to the petal — 0.65+0.45·e−0.85|sin| and 0.68+0.42·e−0.85|sin| — recording the exact moment the load-bearing floor entered the lineage. The floor, not the petal, was the invention. The petal is the ancestor; the axiom is what it became.
Three generations, all recoverable from shipped exports:
14·e^{−0.85·|sin(3θ+0.4l)|}min = e^{−0.85} ≈ 0.427·R₀no designed floor — sinks to decay tail0.65 + 0.45·e^{−0.85|sin|}0.68 + 0.42·e^{−0.85|sin|}floor appears: min 0.842 / 0.8600.55 + 0.45·|sin|^{0.65}·e^{0.10cos}min = 0.55 (attained)floor load-bearingThe signature floor + scale·modulation is visible already in the bridge forms — same structure, un-tuned constants. The petal kept its skeleton; what changed between v1 and v2 is that collapse became structurally impossible rather than merely rare.
| site | form | label in code |
|---|---|---|
| ProceduralGeometry.ts:216 | 14·e^{−0.85|sin(3θ+0.4l)|} α=1 | case 0 — "🌸 Axiomatic Petal" |
| VoxelEngine.ts:342 | baseRad·(0.65+0.45·e^{−0.85|sin|^1.1}) | "Axiom Petal Formula" — bridge |
| VoxelEngine.ts:2369 | 0.68+0.42·e^{−0.85|sin|^1.15} | "Honest Axiomatic Petal" — bridge |
| DiffusionSimulation.ts:549 | e^{−0.85|sin(3θ+0.8t)|^1.4} | "axiomR" — sentinel field |
| BeliefDiffusionView.tsx:332 | same as above | sentinel render |
Two exported builds (NSS Studio, 2026-10-27) — v1 survives in the
petal/sentinel layers while the engine's calculateAxiomRadius runs
v2. One build, two axioms.
| id | claim | how verified | |
|---|---|---|---|
| P-1 | Petal range: [e^{−0.85}, 1]·R₀ = [0.4274, 1]·R₀ — bounded below but only by decay tail, not design | 2×10⁶-point sweep, all α variants | VERIFIED |
| P-2 | 6-fold symmetry: sin(3θ)=0 gives 6 maxima (petal tips) per cycle; the 6 minima sit at θ=π/6+nπ/3 | closed form — extrema of |sin(3θ)| | VERIFIED |
| P-3 | No cusps: α≥1 → folds are smooth (α=1 corner, α>1 differentiable) | |u|^α slope analysis at u=0 | VERIFIED |
| P-4 | Bridge floors: 0.65+0.45·e^{−0.85} = 0.842; 0.68+0.42·e^{−0.85} = 0.860 minimum fractions | closed form | VERIFIED |
| P-5 | The two forms co-exist under one name in shipped builds | grep audit, 2 zip exports | VERIFIED |
The petal proves the axiom wasn't found once — it was arrived at. First a petal that decayed. Then a petal with a floor bolted on (twice, with different constants — the idea was being tuned). Then the floor survived and the petal's decay became the powered fold. The discovery is the arc: nonzero residual went from accident, to bolt-on, to load-bearing theorem.
It also resolves a forensic anomaly: builds labeled "axiom" in four places run v1 while the engine runs v2 — a naming collision, not an error. Both are ours; only one is canon.
Family disclosure: the petal's mechanism — an exponential of a trigonometric kernel — shares its family with the von Mises distribution (e^{κ·cos θ}, the circular Gaussian). The petal uses an |sin(3θ)|^α kernel for 6-fold petals instead of a single concentration; as with the axiom, the ingredient is published and the composition is ours.