Predictions

Reading the tiers honestly. A Tier-A or sub-10-ppm label means "structural derivation present in the kernel chain" — not "agreement at the precision of the experimental measurement." Two Tier-A predictions with very different agreement levels can sit side by side in the catalogue, and a single prediction can carry multiple legitimate derivation routes at different precision bands. For example: the fine-structure constant α is one of the most precisely measured quantities in physics (relative precision ≈ 10⁻¹⁰). Our framework derives α via two routes — a closed-form algebraic LO α = (11/15)²·ι_τ⁴ reproducing CODATA to ~9.8 ppm (auditable in one line; see Book IV Chapter 10), and a full multi-loop derivation (IV.T107 with NLO holonomy + NNLO window algebra) reaching ~0 ppm. Both are honest framework claims; both co-exist by design. The electron mass m_e, derived via the kernel-anchored cascade with the neutron mass as the dimensional anchor, agrees to ~0.025 ppm — and is itself a fit of the dimensionless ratio m_e/m_n, not an absolute prediction. All three are Tier-A. None of them are at measurement precision. Always open the per-prediction page to read the precision band, which derivation route is being cited, the dimensional-anchor chain, and the comparator-vintage details before treating any single tier badge as endorsement of measurement-level agreement. See also question 11 in the Red-team FAQ for the full discussion, including the explicit arithmetic. The predictions browse grid now displays a per-row precision-band chip on every prediction card so the band is readable at a glance.