EVIDENCE_LEDGER.json

{
  "version": "3.0.0",
  "new_laboratory_data": false,
  "entries": [
    {
      "id": "E1",
      "type": "synthetic_manufacturing",
      "count": 512,
      "independence": "32 paired seeds per dimension; eight methods share cases",
      "files": [
        "results/gauge/runs.json",
        "results/gauge/summary.json"
      ],
      "supports": "Behavior of the specified reduced controller model",
      "does_not_establish": "Chemical backend, population yield 1, or superiority to a matched conventional controller"
    },
    {
      "id": "E2",
      "type": "negative_control",
      "count": 36,
      "independence": "Subset of E1, not additional runs",
      "files": [
        "results/gauge/runs.json"
      ],
      "supports": "Differential bias can cause false acceptance despite a small cycle residual",
      "does_not_establish": "Failure probability in an unmeasured physical platform"
    },
    {
      "id": "E3",
      "type": "synthetic_reachability",
      "count": 25000,
      "independence": "Sampled arrays; each reused across 31 capacities",
      "files": [
        "results/gauge/phase.json"
      ],
      "supports": "Agreement with the ideal G4 finite-size curve",
      "does_not_establish": "Universal manufacturability phase transition"
    },
    {
      "id": "E4",
      "type": "numerical_formula_checks",
      "files": [
        "results/gauge/theorem_checks.json",
        "results/gauge/calibration.json"
      ],
      "supports": "100 quadratures, 200 response cases, 6000 covariance draws and three local-solver demonstrations",
      "does_not_establish": "Formal verification or fully local hardware execution of every manufacturing run"
    },
    {
      "id": "E5",
      "type": "automated_scientific_tests",
      "count": 28,
      "files": [
        "tests/test_gauge.py",
        "tests/test_revision.py",
        "results/gauge/test_log.txt"
      ],
      "supports": "Implementation checks, including 80 LP and 500 bounded-noise examples",
      "does_not_establish": "Independent replication or validation of chemical assumptions"
    },
    {
      "id": "E6",
      "type": "preserved_v2_computation",
      "count": 512,
      "files": [
        "results/runs.json",
        "results/V2_RESULTS.md"
      ],
      "supports": "The prior absolute-response contract study under its own model",
      "does_not_establish": "Direct pooled comparison with v3"
    },
    {
      "id": "E7",
      "type": "proposed_experiment",
      "files": [
        "EXPERIMENT_PROTOCOL.md"
      ],
      "supports": "A falsifiable protocol exists",
      "does_not_establish": "An experiment was performed"
    }
  ]
}