Temporal–Generative Alignment Test
Pre-registered, non-tunable experimental design
Mission: Test whether utility emergence requires the temporal alignment of a generative event (γ) with a coherence-stabilized history (κ), rather than static parameters or late optimization.
⚠️ Pre-registration Notice: This chamber implements a pre-registered experimental design. All parameters, hypotheses, success criteria, and falsifiers are locked before execution. No post-hoc tuning is permitted. Only γ_injection_time varies; all other parameters are frozen.
🔒 Frozen Parameters
Grammar
axis_ii_fixed_v1
Topology
XLIV / XLV reference topology
Depth
400 steps
Utility Definition
G_degree === 1 (unchanged)
ε (Asymmetry)
0.10 (fixed)
κ (Coherence)
4 (fixed)
Seeds per Point
20 (fixed)
🎯 Experimental Design
H₃₁: Temporal–Generative Alignment
Utility emerges only when γ is injected during a coherence-stabilized history window.
G° ≠ f(ε)
G° ≠ f(κ)
G° ≠ f(γ)
G° = f(history alignment of κ and γ)
✓ ∃ t: Ufreq(t) ≥ 0.25
✓ Utility appears only in bounded temporal window
✓ No utility at γ_time = 0 or very late
✓ Reproducible across seeds (low variance)
📋 Sweep Results
γ_time
Ufreq
Pcollapse
Lpersist
Uonset
Seeds
Status
No data yet. Run experiment to begin.
✅ Acceptance Criteria
Chamber XLVI is SUPPORTED if and only if:
∃ t : Ufreq(t) ≥ 0.25
Utility appears only in contiguous time window
Utility does not appear when γ_time = 0
Utility does not appear when γ_time ≫ stabilization
Results reproduce across seeds (low variance)
❌ Falsification Criteria
Chamber XLVI is FAILED if any of:
F1 — Generative Dominance:
Ufreq > 0 for all γ_injection_time
→ Axis II (operator-level control of utility) is falsified under preregistered tests. Utility does not emerge from operator timing or alignment alone.
F2 — Coherence Irrelevance:
Ufreq > 0 when κ = 0
→ Axis II (operator-level control of utility) is falsified under preregistered tests. Utility does not emerge from operator timing or alignment alone.
F3 — No Emergence:
Ufreq = 0 for all γ_injection_time
→ Axis II (operator-level control of utility) is falsified under preregistered tests. Utility does not emerge from operator timing or alignment alone.