V2.389 - Species-Dependence Falsification Map — 5 Unique Testable Predictions
V2.389: Species-Dependence Falsification Map — 5 Unique Testable Predictions
Question
What predictions does this framework make that no other dark energy model shares? For each, what exact measurement would confirm or kill it, by which experiment, and when?
The 5 Unique Predictions
Every other dark energy model (ΛCDM, quintessence, modified gravity) treats Λ as independent of particle physics. This framework says Λ = f(field content). Five consequences follow that no competitor shares:
| # | Prediction | Unique because | Kill experiment |
|---|---|---|---|
| 1 | Λ shifts calculably with new particles | ΛCDM: Λ is free | LHC + Euclid |
| 2 | N_eff = 3.044 required for correct Λ | ΛCDM: N_eff independent of Λ | CMB-S4 (2030) |
| 3 | Majorana neutrinos required | ΛCDM: agnostic | LEGEND/nEXO (2028-32) |
| 4 | BH log correction γ = -149/12 | LQG: γ = -3/2 (universal) | QG phenomenology |
| 5 | m_graviton = 0 exactly | dRGT: m_g ~ H₀ explains DE | LISA (2035+) |
Per-Particle Sensitivity
| BSM particle | ΔR | σ (Planck) | σ (Euclid) | Status |
|---|---|---|---|---|
| Real scalar (axion) | -0.005 | -0.6 | -2.4 | allowed → marginal |
| Complex scalar | -0.009 | -1.3 | -4.7 | allowed → disfavored |
| Weyl fermion (sterile ν) | -0.007 | -1.0 | -3.6 | allowed → disfavored |
| Dirac fermion | -0.014 | -2.0 | -7.1 | allowed → EXCLUDED |
| Vector boson (Z’) | +0.027 | +3.7 | +13.5 | disfavored → EXCLUDED |
| 2nd Higgs doublet | -0.019 | -2.5 | -9.2 | marginal → EXCLUDED |
| 4th generation | -0.102 | -14.0 | -51.1 | EXCLUDED |
| MSSM | -0.251 | -34.3 | -125.4 | EXCLUDED |
| Dark SU(2) | +0.079 | +10.8 | +39.3 | EXCLUDED |
| Dark SU(3) | +0.195 | +26.7 | +97.5 | EXCLUDED |
The direction of the Λ shift identifies the spin of the new particle. Scalars/fermions decrease Λ; vectors increase it. This is testable.
Species-Dependence Table
| Scenario | R | Λ/Λ_obs | σ (Planck) | σ (Euclid) |
|---|---|---|---|---|
| SM + graviton | 0.688 | 1.004 | +0.4 | +1.5 |
| + 1 axion | 0.683 | 0.998 | -0.2 | -0.8 |
| + 1 sterile ν | 0.681 | 0.994 | -0.6 | -2.1 |
| + 3 right-handed ν (Dirac) | 0.667 | 0.974 | -2.5 | -9.0 |
| + 1 dark photon | 0.715 | 1.044 | +4.1 | +15.0 |
| + 2HDM | 0.669 | 0.978 | -2.1 | -7.7 |
| + 4th generation | 0.586 | 0.855 | -13.6 | -49.6 |
| + MSSM | 0.437 | 0.638 | -33.9 | -123.9 |
Experimental Timeline
| Year | Experiment | Framework prediction | Kill threshold |
|---|---|---|---|
| 2026 | DESI DR2 | w₀ = -1.00 | Current 3.9σ tension (SNe-driven) |
| 2027 | DESI DR3 | w₀ = -1.00 ± 0.00 | 5σ if w₀ = -0.85 persists |
| 2028 | Euclid | Ω_Λ = 0.688 ± 0.002 | 3σ if outside [0.682, 0.694] |
| 2028 | LEGEND-200 | Majorana neutrinos | Dirac → R shifts by -0.02 |
| 2030 | CMB-S4 | N_eff = 3.044 ± 0.06 | 17σ if N_eff = 4 |
| 2035 | LISA | Standard EW spectrum | Anomalous → new EW physics |
| 2040 | Cosmic variance | Ω_Λ at σ = 0.001 | Pins n_grav to 10 or 11 |
The Acid Test
If a new light particle is discovered at ANY collider, the framework makes an IMMEDIATE prediction for how Ω_Λ must shift. If Euclid confirms the shift: framework wins. If Euclid sees no shift: framework is dead.
No other framework can be killed this way. This is the single most powerful distinguishing feature.
Most Vulnerable Point
DESI DR2’s w₀ = -0.75 ± 0.06 is a 3.9σ tension with the framework’s prediction w₀ = -1. However, V2.377 showed this tension is entirely SNe-driven — DESI’s own BAO data prefers w₀ = -1. If DESI DR3 (2027) confirms w₀ ≠ -1 at >5σ with independent calibration, the framework is falsified.
What This Means
This experiment assembles the framework’s complete falsification map. The framework makes 10+ quantitative predictions, each independently testable within 5-15 years. Five of these are UNIQUE — not shared with any competitor. The framework can die at least 5 independent ways. That is what makes it physics.
Files
src/falsification_map.py: All predictions, sensitivities, timelinetests/test_falsification_map.py: 19 tests, all passingrun_experiment.py: Full 8-section analysisresults.json: Machine-readable output