Documentation

CsdLean4.RecordLayer.RecordLayerClosure

SigmaLayer/RecordLayerClosure: the record-layer capstone bundle (MD-1, step 5) #

Category: 7-SigmaLayer (the record layer — the closure-level statement).

The record-layer analog of SigmaLayer/FiniteQMClosure.lean: a single Prop bundle RecordLayerClosure collecting exactly the record-layer facts that are GENUINELY PROVED for the fibre-partition readout, discharged in one theorem recordLayerClosure. This is the successor readout to the preparation-indexed LF5/PointerOutcome.lean (vnPointerOutcome) that the closure currently carries: unlike that engine, here the outcome probabilities are measurement-noncontextual (the Born weight of outcome i is ‖ψ i‖², depending only on (eᵢ, ψ)), while the record events are a genuine postulate-P5 RecordSemantics.

The bundle certifies, on the fibre Σ = ℝ with the canonical typicality measure fibreTypicality, for the Born context of a unit state ψ:

⚠️ SUPERSEDED 2026-07-31 by SigmaLayer/GlobalRecordClosure.lean #

This bundle remains true and is not deprecated, but it is no longer the record layer's best statement. Two defects, both fixed by the successor:

The five closure fields are otherwise identical, which is the evidence that neither defect was ever load-bearing for the record layer's content. ae_total also strengthens — though ⚠️ not for the reason this file used to give. It said the form "must be stated relative to Ico 0 1 because Lebesgue measure on the line is infinite". fibreTypicality is a probability measure, not Lebesgue on the line, and fibreTypicality_uncovered_univ proves the univ form on outright. The genuine strengthening is that there the mass one is Haar mass on a compact space, whereas here it is imposed by restriction and the fibre's complement is excused rather than covered (fibreTypicality_Ici_one).

Kept rather than deleted because it is consumed and true. New work should cite the successor.

Honest scope (what step 5 does and does not do) #

This is the record-layer readout as a first-class, certified bundle — the replacement candidate for vnPointerOutcome. It does not swap the readout inside unifiedFiniteQMClosure's proved fields: those are theorems on the productDynamics engine over ℂℙ^M × T² (bornRegion/vnPointerOutcome), and re-deriving them on this fibre model is the open engine-migration work, not a mechanical edit. Likewise the physical de-isolation flow generating the cells remains open (plan §3c / step 2b′). So step 5 lands the successor and its closure-level statement; the full migration of FiniteQMClosure onto it, and the flow behind it, stay open (MD-1). Foundational-triple, no sorry.

References #

specs/record-layer-plan.md (record layer, MD-1; step 5); SigmaLayer/FibreRecord.lean (the RecordSemantics instance); SigmaLayer/DeIsolationFlow.lean (fibreTypicality, fibreTypicality_uncovered); SigmaLayer/FiniteQMClosure.lean (the closure this parallels, and whose vnPointerOutcome readout this succeeds).

The record-layer closure bundle (MD-1). The record-layer facts genuinely proved for the fibre-partition readout of a unit state ψ, collected into one Prop and discharged by recordLayerClosure. The successor to the preparation-indexed vnPointerOutcome readout: the outcome probabilities are measurement-noncontextual (born_typicality) and the records are a first-class postulate-P5 RecordSemantics.

Instances For

    The record-layer closure holds for every unit state. Each field is discharged by its source lemma in FibreRecord.lean / DeIsolationFlow.lean — the record-layer readout is a certified bundle, the successor to the preparation-indexed vnPointerOutcome.