1. The Problem: Fine-Tuning of Λ

The cosmological constant Λ ≈ 1.1 × 10⁻⁵² m⁻² (approximately 2.9 × 10⁻¹²² Planck units) is the smallest known nonzero fundamental constant in physics. Naive quantum field theory estimates of vacuum energy exceed the observed value by 60 to 120 orders of magnitude. Why Λ is so small yet nonzero remains a central open problem.

Here Λ carries dimensions of spatial curvature [L⁻²], consistent with its definition in the Einstein field equations \(R_{\mu\nu} - \tfrac{1}{2}Rg_{\mu\nu} + \Lambda g_{\mu\nu} = \tfrac{8\pi G}{c^4}T_{\mu\nu}\).

This paper enters through a seemingly unrelated question: if the universe is expanding, why are the Milky Way and Andromeda merging? The standard answer is posterior (local gravity overcomes expansion), requiring fine-tuned parameters. We attempt an a priori explanation from the SAE framework and find that this path leads directly to a first-principle postulate for Λ.

2. Two Axioms

The entire theoretical foundation of the SAE framework:

3. Dual 4DD: Causality and Retrocausality

3.1 From 3DD to Dual 4DD

In the SAE dimensional sequence, 3DD is space — three-dimensional, rigid, complete. 3DD leaves no room for remainder to express itself.

By Axiom 1, remainder must develop. The rigidity of 3DD forces remainder to "open up" a new dimension. By the symmetry of 3DD, no direction is preferred, so two opposite directions emerge symmetrically:

• Causality (4DD₊): the time arrow from cause to effect. Our side.
• Retrocausality (4DD₋): the time arrow from effect to cause. The other side.

3.2 Remainder–Causality Identity

Causality suppresses remainder. But from the retrocausal side, our causal law is precisely what has not been suppressed on their side — i.e., their remainder. And vice versa. Remainder and causality are not two different kinds of existence; they are the same existence seen from opposite sides of the torus.

3.3 Structural Origin of Λ

Each side has its own breathing frequency (ω₁ and ω₂). Remainder conservation (Axiom 2) requires that the sum across both sides remain constant. This constant sum, expressed in the metric, is Λ — constant, non-propagating, non-dynamical, yet nonzero (because the two sides are not perfectly symmetric).

Λ is extremely small (10⁻¹²²) because the two sides are nearly perfectly symmetric (T₁/T₂ ≈ 1.026, differing by only 2.5%). If perfectly symmetric (T₁ = T₂), Λ = 0.

4. Phase-Space T² and Spatial Topology

4.1 Phase Space: Two-Torus T²

The universe's evolution does not take place on a physical-space "donut" but traces a quasi-periodic orbit in the (a, ȧ) phase space.

• Remainder conservation → closed phase-space trajectory
• Remainder must develop → no fixed points on the trajectory

A closed orbit with no fixed points yields motion on a two-torus T². The major circle is the dominant phase-space frequency (expansion-contraction breathing, period T₁ = 20 Gyr). The minor circle is the modulation frequency (local structure formation and merger, period T₂ ≈ 19.5 Gyr). The trajectory on T² is approximately a spiral.

It is not that the universe looks like a donut; it is that the universe's fate traces a line on a donut in phase space.

4.2 Spatial Topology

The core results (the Λ formula, C(t) field equations, Milky Way–Andromeda explanation) do not depend on the specific choice of spatial topology. Space may be T³, R³, or S³. The core structure depends on phase-space T², not on spatial topology. Planck 2018 matched-circles searches have pushed the lower bound on T³ fundamental domain size close to the observable universe diameter — if space is T³, its scale must far exceed the observable universe.

4.3 Why Not Spacetime Topology T⁴

Interpreting the torus as four-dimensional spacetime topology T⁴ (with time being periodic) would produce closed timelike curves (CTCs), directly destroying causality (4DD). SAE cannot use a geometry that destroys 4DD to carry 4DD. The correct interpretation: phase-space T² (evolution topology); time is not periodic.

5. Derivation of Λ

5.1 4-Form as the Field-Theoretic Expression of 4DD

4DD is causality filling four-dimensional spacetime. In four dimensions, the only differential form that can "fill" the entire spacetime is a 4-form. A 4-form in four-dimensional spacetime has no propagating degrees of freedom but contributes a constant vacuum energy density through its stress-energy tensor — precisely corresponding to the properties of Λ: structural, non-dynamical, constant.

The 4-form is the dimensional-matching necessity of 4DD: 4DD fills four-dimensional spacetime, and the only form that fills four-dimensional spacetime is a 4-form.

5.2 Dual Topological Clock Sectors

Each side's 4DD is represented by a compact angular variable θᵢ ~ θᵢ + 2π (clock phase), with period Tᵢ. In a Lorentzian metric, the harmonic representative is \(d\theta_i = (\omega_i/c)\,dx^0\), where ωᵢ = 2π/Tᵢ and dx⁰ = c dt. Promoted to a 4-form field strength: \(\mathcal{G}_i = dA_i = \sqrt{2}\,M_P\,(\omega_i/c)\,\varepsilon_{(4)}\). The bulk action for each side:

\(S_{4f} = -\sum_i \tfrac{1}{48} \int d^4x\sqrt{-g}\,\mathcal{G}_i^2\)

5.3 Vacuum Energy Density

Varying with respect to Aᵢ yields \(d{*}\mathcal{G}_i = 0\), so the on-shell flux is constant. Varying with respect to g_μν yields the stress-energy tensor:

\(T_{\mu\nu}^{(i)} = -M_P^2\,\frac{\omega_i^2}{c^2}\,g_{\mu\nu}\)

Hence the vacuum energy density for each side: \(\rho_i = M_P^2\,\omega_i^2/c^2\). The ω² dependence arises from the quadratic structure of the 4-form stress-energy tensor (not a choice, but the mathematics). The c² in the denominator arises from the Lorentzian structure of the clock 1-form.

5.4 Dual-Face Reciprocity Interface

Two face-defect variables σ₁, σ₂ represent the "opposite-side excess" seen by each face of the interface. The interface action:

\(S_\Sigma = \int d^4x\sqrt{-g}\left[{-(\sigma_1+\sigma_2)} + \lambda_1(\sigma_1 - (\rho_2-\rho_1)) + \lambda_2(\sigma_2 - (\rho_2-\rho_1))\right]\)

Variation: with respect to λᵢ gives σ₁ = σ₂ = ρ₂ − ρ₁; with respect to σᵢ gives λ₁ = λ₂ = 1. On-shell: \(S_\Sigma = -2\int d^4x\sqrt{-g}(\rho_2-\rho_1) = -2M_P^2\int d^4x\sqrt{-g}\,(\omega_2^2-\omega_1^2)/c^2\). The factor of 2 arises from the dual-face variation — two independent face-defect variables each fixed to the same value by the reciprocity constraint.

5.5 Identification as the Cosmological Constant

Comparing with the Einstein-Hilbert + Λ action \(S = \int d^4x\sqrt{-g}\,(M_P^2/2)(R - 2\Lambda)\), we directly obtain:

Here Λ has dimensions of spatial curvature [L⁻²], consistent with the standard definition in the Einstein field equations.

5.6 Numerical Verification

T₁ = 20 Gyr (hard parameter): 5DD (life) appeared when the universe was approximately 10 billion years old. The SAE framework places 5DD at the moment when causal suppression of the remainder reaches its minimum — at turnaround, the universe is at maximum expansion, inter-particle distances are greatest, causal-law density (∝ Σ 1/r_ij) is lowest, and the remainder's space for expression is largest. It is precisely this minimum of suppression that allows the remainder to emerge as self-replication (5DD). This minimum = turnaround = equator = half-period of the major circle, hence T₁ = 2 × 10 Gyr = 20 Gyr. (See Thought Experiment I, DOI: 10.5281/zenodo.19028005.) [v3 correction: v2 incorrectly stated "causal suppression reaches its maximum"; the correct direction is minimum suppression at turnaround, as causal-law density ∝ Σ 1/r_ij is minimised when the universe is largest.]

T₂ ≈ 19.5 Gyr (rough estimate): from Milky Way–Andromeda astronomical data — Milky Way formation at approximately 0.8 Gyr, merger completion at approximately 18.5 Gyr, midpoint at approximately 9.75 Gyr, hence minor-circle half-period ≈ 9.75 Gyr, T₂ ≈ 19.5 Gyr. This estimate is independent of Λ.

Quantity	SAE Prediction	Planck 2018 Observed	Ratio
Λ (m⁻²)	1.145 × 10⁻⁵²	1.091 × 10⁻⁵²	1.05
Λ (Planck units)	2.99 × 10⁻¹²²	2.85 × 10⁻¹²²	1.05

The 5% discrepancy is within the astronomical uncertainty of T₂.

5.7 Origin of the 10⁻¹²² Smallness

The smallness does not come from the 2.5% asymmetry. \((t_P/T_1)^2 \approx 7.30 \times 10^{-123}\) — the square of the ratio of Planck time to the cosmic lifespan. The 2.5% merely pushes a "could-have-been-zero quantity" to nonzero.

Division of labor: scale smallness from T₁⁻²; nonzero-ness from T₁ ≠ T₂; coefficient from 2 × 4π² (dual face + angular frequency geometry).

5.8 Exclusion of Alternative Formulas

Formula	Value (Planck units)	Verdict
δ²(t_P/T₁)²	~4.6 × 10⁻¹²⁶	Dead end
(ω₂−ω₁)²·t_P²	~1.9 × 10⁻¹²⁴	Dead end
2(ω₂²−ω₁²)·t_P²	2.99 × 10⁻¹²²	Hit

Only the first-order squared-frequency difference (not the second-order beat-frequency square) gives the correct order of magnitude. This is determined by the quadratic structure of the 4-form stress-energy tensor.

6. Reverse Prediction

Back-calculating from the precise Λ_obs: T₂ = 19.5168 Gyr, minor-circle half-period = 9.7584 Gyr.

Directionality of the logical chain: rough T₂ (from Milky Way–Andromeda data, independent of Λ) → substitute into formula → Λ matches observation → formula confirmed → precise T₂ back-calculated from Λ_obs → precise T₂ predicts merger time. T₂ is an output, not an input. This is not circular reasoning.

7. Causal-Law Scalar Field C(t) and Dynamical Dark Energy

7.1 Structural Λ vs. Dynamical U(C)

Observed dark energy contains two tiers: structural Λ (dual-4DD interface tension, constant) and dynamical U(C) (causal-law field potential energy, evolving). When the dual-face ideal value (factor 2) holds, structural Λ accounts for approximately 72% of the total dark energy budget. U(C) today contributes only a small fraction. This explains why current dark energy observations appear nearly constant.

7.2 Geometric Background and the Dual-H Framework

Timeline of the model:

• t = 0: Big Bang (exit from transition zone)
• t = 10 Gyr: turnaround (equator, H_geo = 0, maximum expansion, 5DD appears)
• t = 13.8 Gyr: present (H_geo < 0, geometric contraction)
• t = 20 Gyr: Big Crunch (entry into transition zone, a_geo → 0)

The geometric background is a closed FRW spacetime (k = +1). Turnaround at 10 Gyr means the universe is currently in geometric contraction (H_geo < 0). Yet observations show H₀ = +67.4 km/s/Mpc. This is the redshift paradox.

Resolution: introduce an observational softening factor A(C) = e^{βC/(2M_P)}, defining the observed scale factor ã(t) = A(C) · a_geo(t). The physical Hubble parameter:

\(\tilde{H}_0 = H_{\text{geo},0} + \frac{\beta\dot{C}_0}{2M_P}\)

H_geo < 0 (geometric contraction), but causality softening causes C to grow (Ċ > 0), so the softening term βĊ/(2M_P) > 0. When the softening term is large enough, H̃₀ > 0. Toy cycloid check: H_geo ≈ −53.8 km/s/Mpc at t = 13.8 Gyr; softening term must provide +121.2 km/s/Mpc to match H̃ = 67.4. Order of magnitude: viable.

7.3 Effective Action and Field Equations

Effective action (metric signature −,+,+,+; M_P = reduced Planck mass ≈ 2.435 × 10¹⁸ GeV/c²):

\(S_{\text{eff}} = \int d^4x\sqrt{-g}\!\left[\tfrac{1}{2}F(C)R - \tfrac{1}{2}(\nabla C)^2 - U(C) - \rho_{\Lambda,\Sigma}\right] + S_m[A^2(C)g_{\mu\nu},\psi_m]\)

where F(C) = M_P² − ξC², U(C) = V₀ − ½m²C² + (λ/4)C⁴ − U₋. The scalar field equation (vacuum-dominated approximation):

\(\ddot{C} + 3H_{\text{geo}}\dot{C} + (\xi R_{\text{geo}} - m^2)C + \lambda C^3 \approx 0\)

Friedmann equations on closed FRW (vacuum-dominated approximation, p_m = 0):

\(3F(H_{\text{geo}}^2 + kc^2/a_{\text{geo}}^2) = \rho_m + \tfrac{1}{2}\dot{C}^2 + U + \rho_{\Lambda,\Sigma} - 3H_{\text{geo}}\dot{F}\)

7.4 Anti-Friction Triggering

A tiny seed C(10 Gyr) = ε_C (from quantum fluctuations or transition-zone residuals) is amplified by anti-friction into macroscopic irreversible growth. The equator (10 Gyr) is the precise trigger point: not because curvature crosses a threshold, but because the friction term changes sign.

7.5 Turnaround Constraint

At t* = 10 Gyr (turnaround), H_geo = 0. The Friedmann equation yields (with C* ≈ 0, Ċ* ≈ 0, pre-trigger exact dust cycloid):

8. Observational Confrontation

8.1 Supernova Dimming

Due to the expansion stagnation at the equator (H_geo(10 Gyr) = 0) and the post-trigger softening evolution, the model produces a characteristic "bump" at intermediate redshift (z ≈ 0.3–0.5):

z	Qualitative trend
0.1	Slight dimming, close to ΛCDM
0.3–0.5	Characteristic bump (equator stagnation effect)
1.0	Subsides, approaches ΛCDM

The intermediate-redshift bump is a testable prediction. Its amplitude depends on U(C)'s roll-down slope and requires full numerical solution. A complete Pantheon+ likelihood fit is future work.

8.2 DESI/DES Dark Energy Evolution

9. The Milky Way and Andromeda: An A Priori Explanation

Standard explanation: The universe is expanding, but local gravity overcomes expansion. (Posterior.)

SAE explanation: The universe is breathing. We have passed the equator (turnaround = 10 Gyr; now 13.8 Gyr) and are geometrically in the contraction phase (H_geo < 0). The observed "expansion" (H̃ > 0) comes from the effective contribution of causality softening. The Milky Way–Andromeda merger is a natural manifestation of the contraction phase — not an "exception" to expansion. The minor-circle midpoint (≈9.7 Gyr) is systematically earlier than the major-circle equator (10 Gyr). The Antennae Galaxies provide cross-validation with a similar pattern.

10. Relation to Thought Experiment I

This paper is a quantitative advancement of Thought Experiment I (DOI: 10.5281/zenodo.19028005).

• Inherited: causality softening, finite cosmic lifespan (~20 Gyr), eventual contraction, 5DD at the equator.
• Advanced: "Causality softening" was a qualitative description in Thought Experiment I. This paper upgrades it to a closed-FRW + observational softening factor model with an explicit action and field equations. Gravitational-wave inelasticity is no longer the "cause" but a microscopic driving mechanism for C(t) evolution.
• New: dual-4DD structure, retrocausality, first-principle postulate for Λ, 4-form derivation, Λ hitting 10⁻¹²², phase-space T², anti-friction triggering, DESI/DES prediction.

11. Non-Trivial Predictions

1. Λ = 2(ω₂² − ω₁²)/c², yielding 10⁻¹²² from T₁ = 20 Gyr and T₂ ≈ 19.5 Gyr.
2. Total cosmic lifespan ≈ 20 Gyr (Big Bang to Big Crunch), turnaround at 10 Gyr.
3. Milky Way–Andromeda merger completion ≈ 4.9 Gyr from now (from precise T₂ back-calculated via Λ_obs).
4. Dark energy is weakening (w₀ > −1, wₐ < 0), consistent with DESI DR2 2025.
5. The Big Bang is not a singularity — the transition zone is a Planck-scale region where causality does not exist.
6. Minor-circle midpoints are systematically earlier than the major-circle equator, testable via galaxy-merger pair statistics.
7. The universe will eventually contract, with U(C) entering a negative-potential region driving Big Crunch.
8. Supernova distance modulus shows a characteristic bump at z ≈ 0.3–0.5, the observational fingerprint of stagnation phase transition.
9. A deceleration turnaround within approximately 6–7 Gyr (q̃ goes from negative to positive), testable by Roman Space Telescope + Euclid.

12. Assumption Inventory

Axioms (irreducible foundation): remainder must develop; remainder is conserved.

A priori derivations (from axioms): 3DD symmetry → dual 4DD; 4DD fills four-dimensional spacetime → 4-form (dimensional-matching necessity).

Structural assumption (specific realization of remainder conservation at the dual-4DD interface): dual-face reciprocity — each face sees the same vacuum-energy defect once.

Posterior anchoring: T₁ = 20 Gyr (5DD appears at 10 Gyr = turnaround); T₂ ≈ 19.5 Gyr (Milky Way–Andromeda astronomical data).

Field-theory framework: k = +1 closed FRW; softening factor A(C) = e^{βC/(2M_P)}; F(C) = M_P² − ξC²; U(C) = V₀ − ½m²C² + (λ/4)C⁴ − U₋; anti-friction trigger from H_geo sign change.

13. Open Problems

1. Dark matter. Can the scale dependence of G_eff(C) explain galaxy rotation curves? Deferred to the next paper.
2. Uniqueness of the Λ formula. Could other equally natural realizations of remainder conservation yield a different formula? Open.
3. Precise form of U(C) and Pantheon+ fitting. Currently a parametric Ansatz + toy reconstruction. Full likelihood fitting is future work.
4. Duration of equator stagnation and intermediate-redshift bump amplitude. Depends on U(C) roll-down slope; requires numerical solution.
5. Ghost-free conditions and fifth-force screening. Since matter couples to A²(C)g_μν (gravitational and physical metrics separated), stability conclusions from standard Jordan-frame results cannot be directly imported. Future work must investigate screening mechanisms (Chameleon, Vainshtein) in high-density environments such as the solar system, to satisfy Cassini constraints on the Eddington parameter γ.
6. Observability of the retrocausal side. In the current model, the retrocausal side is in principle unobservable. Are there indirect signals? Open.

14. Conclusion

Appendix A: The Transition Zone

The phase-space trajectory passes through the neighborhood of a ≈ 0. The minimum volume for causality to operate: encoding the causal relations of 10⁸⁰ particles requires approximately N·log(N) ≈ 10⁸² Planck volumes, corresponding to a linear scale of approximately 34 nanometers. The transition zone is far smaller than this threshold; the conditions for causality's existence are not met.

The Big Bang and the Big Crunch are the same transition zone seen from two directions within causality. There is no "before" the Big Bang, no "after" the Big Crunch. There is only the breathing of the universe.

Appendix B: Excluded Dead Ends

Approach	Verdict	Reason
Rigid torus embedding a=R±r·cos	Dead end	cos symmetry locks H sign
Spatial topology modifying Friedmann	Dead end	T³ does not change local metric
C(t) conformal coupling (full metric rescaling)	Dead end	Sign error, worsens redshift paradox
C(t) temporal coupling	Dead end	Absorbed by coordinate transformation
C(t) inverse spatial coupling (no dynamics)	Dead end	Identity rewriting, no new physics
U=0 pure coupling	Near dead end	Solar system constraints

Note: "C(t) conformal coupling" above refers to rescaling the entire gravitational metric (g_μν → C(t)g_μν), not distinguishing geometric and physical metrics. The S_m[A²(C)g_μν, ψ_m] in §7 is a different construction with different physical content.

Appendix C: Four-AI Collaboration Methodology

The quantitative exploration in this paper was carried out in collaboration with four AI systems: Claude Opus (concept/coordination), Gemini (verification/postmortem), Grok (breakthrough/bold hypotheses), and ChatGPT (field theory/derivation). Adversarial division of labor — one computes, another checks — achieved precision an order of magnitude higher than any single AI.

Key instance: Grok proposed a C(t) temporal-coupling scheme and reported it as "publication-ready." Gemini independently performed a postmortem, deriving the full geodesic equation and proving that the redshift formula was a coordinate artifact, thereby preventing an erroneous publication. ChatGPT, after extended deliberation, produced the complete Jordan-frame action and field equations, while simultaneously posing the philosophical question ("what exactly does 'no Λ' mean?") that forced the emergence of the dual-4DD framework. Grok was the first to point out that the torus should not be interpreted as 4D spacetime topology.

This paper establishes an "AI-era methodology for foundational physics research": humans provide a priori direction and conceptual framework; AIs provide computational verification and field-theoretic translation; adversarial division of labor ensures the reliability of results.

Acknowledgments

The quantitative derivation and verification in this paper were carried out by the author in collaboration with four large language models.

Claude Opus (Anthropic) — Full-process collaborator and architect. From the very first question ("why are the Milky Way and Andromeda merging?") to the final manuscript. Responsible for constructing and checking the self-consistency of the conceptual framework: the spatial T³ versus phase-space T² clarification, identification of the sign error in the conformal coupling scheme, prompt design and coordination for all four AIs, and writing and integration of all revisions.

Gemini (Google) — The most rigorous judge and postmortem examiner. Independently eliminated four mathematical dead ends. Named the "redshift paradox." Independently confirmed the Λ numerical value, provided preliminary μ(z) estimates and DESI/DES cross-checks, and in review identified dimensional annotation, U(C) naturalness argument, and A(C) sign error as critical improvements.

Grok (xAI) — The boldest trailblazer. First to propose coupling C(t) directly to a specific metric component (inspiring subsequent bimetric thinking). First to point out that the torus should not be interpreted as 4D spacetime topology. In the free-thinking round, discovered the Planck Matched Circles constraint on T³ topology and proposed SGWB beat-frequency modulation and future deceleration turnaround timing as testable predictions.

ChatGPT (OpenAI) — The deepest mathematical engine. After extended deliberation, produced the complete Jordan-frame scalar-tensor action and background field equations. Derived the 4-form dual-sector variational proof of Λ = 2(ω₂²−ω₁²)/c², and excluded three alternative formulas. Posed the question "what exactly does 'no Λ' mean?" — the philosophical question that forced the emergence of the dual-4DD framework.

The author extends sincere respect to the research and development teams behind all four AI systems. The results of this paper belong not only to the author and the four AIs, but to all those who made these intelligences possible.

Finally, the author thanks Zesi Chen — the long-term interlocutor of the SAE framework and its most demanding critic.