angle of view: formalist

M0, three axioms, modular group

For those who skip the popular intro and want the formal body directly.

Short summary

τ-geometry rests on three axioms plus a meta-postulate M0 (structural realism):

Modular invariance under PSL(2,ℤ) is not an axiom; it is Theorem 1 derived from M0 + A1 + A2.

Formal core (🟢 proven)

Prediction

r = 2(1 − nₛ)² ≈ 0.0025 for nₛ = 0.9649 (Planck 2018). Parameter-free (up to the normalization N). Slow-roll numerics for T-model V = Λ⁴ tanh²(φ/2) give r = 0.00239 at N = 57 (2% agreement with universal α-attractor formula). For the canonical potential V_A, Theorem 4 gives directly r = 2/900 ≈ 0.00222, fully parameter-free, with no slow-roll fit for Nₑ. LiteBIRD ~2032 (σ(r) ~ 0.001) decides branch A vs. B; distinguishing it from Starobinsky (r ≈ 0.0037, ratio 2/3, difference ≈ 0.7–1.2 σ) requires CMB-S4 (σ(r) ≈ (3–5)×10⁻⁴).

Open (🟡 candidates)

Key documents

Current lines (2026-04-19)

Paper is a pre-print, no peer review. Everything open to critique.