classifyThreshold_hypothesis
plain-language theorem explainer
Packages the calibrated classification threshold 0.9 as a quarantined data hypothesis, with provenance, status, and an explicit falsifier. Verification authors cite it when bridging empirical calibration into claim-level interfaces without importing raw measurements onto the certified surface. The body is a thin constructor that lifts the raw MeasurementResult via the standard from-measurement bridge.
Claim. The classification-threshold hypothesis records the claim that the optimal exact-token-match threshold is $0.9$ (uncertainty $0.05$), as a data-derived hypothesis on $\mathbb{R}$. It is falsified if a different threshold achieves better accuracy, and is marked for eventual derivation from stability analysis and information theory.
background
This module is the quarantine layer for empirical inputs. Certified theorem-level claims must not import raw measurements; instead, values enter through provenance records, measurement wrappers, and hypothesis bridges. Modules under Verification/Measurement stay quarantined; only preregistered tests may mix them with the certified surface.
A data hypothesis carries a name, a claimed value of type $\alpha$, the underlying measurement (value, optional uncertainty, provenance), a status flag, and a falsification condition. The raw classification threshold is a calibration constant, not a derived RS quantity: value $0.9$, uncertainty $0.05$, provenance tagged as calibrated for exact token match (dated 2026-01-06).
The local purpose is bookkeeping: keep empirical knobs visible and auditable so they cannot silently underwrite forcing-chain or constant claims (T5–T8, $\alpha$ band, mass ladder) on the certified surface.
proof idea
Not a proof: a definitional wrapper. It applies the from-measurement constructor on the data-hypothesis structure, supplying the name string, the raw measurement record (value $0.9$), the falsifier string ("Different threshold achieves better accuracy"), and the intended derivation path ("Derive from stability analysis + information theory"). No lemmas are invoked; fields are filled by construction.
why it matters
Keeps the classification threshold off the certified surface while still naming it as a first-class hypothesis with a concrete falsifier and a stated closure path (stability analysis plus information theory). Downstream use count is currently zero; the declaration exists so future audited classification interfaces can depend on a hypothesis object rather than a bare float.
It does not touch the forcing chain (T5 J-uniqueness, T6 $\varphi$, T7 eight-tick, T8 $D=3$) or RS-native constants. Its role is methodological: enforce the quarantine rule that measurement knobs remain explicit until replaced by derived claims. The open task it points at is deriving the $0.9$ threshold (or a replacement) from stability and information-theoretic bounds rather than calibration.
Switch to Lean above to see the machine-checked source, dependencies, and usage graph.