Predictions & Falsification

Accountability, Not Decoration

Predictions and falsification are not merely outputs of the program. They are accountability structures for Results.

Predictions and falsification are Verify-owned accountability surfaces.

The Results lane provides world-facing interpretation and result-family context. Verify owns the accountability grammar: prediction IDs, target results, failure conditions, timing assumptions, falsification paths, and pack structure.

Publication artifacts such as the numerical prediction supplement may package these surfaces for reading and review.

Falsification inside the verification matrix

↓ Download print master · 1536 × 864 JPG · CC BY 4.0

Predictions and falsification are inspection routes for empirical claims. They complement formal proof checking and bridge adequacy; they do not replace them.

Lean linkage on every prediction

Every public prediction now carries an auto-derived Lean-modules linkage: the set of TauLib modules in its registry-depends_on closure with lean.formalization in {formalized, skeleton}. The linkage is visible directly on each prediction’s detail page (e.g. Three generations of fermions) as a chip-row of GitHub-linked module names.

From Kernel to Measurement

↓ Download print master · 1536 × 864 JPG · CC BY 4.0

The theory distinguishes internal construction from empirical exposure. Formal derivation becomes physics only through measurement bridges, observable translation, predictions, and falsification pressure.

Prediction and falsification surfaces are downstream of the Construction Spine’s Measurement Bridges step. Internal physical grammar is not empirical accountability until the bridge and failure condition are exposed.

Cross-Lane Reading

The Results lane presents what the program currently derives. This Verify subtree asks how those results could be checked or defeated. For physics, the first concrete surfaces are the Predictions browse, the Prediction Timing Ledger, and the Falsification Pack.