{
  "idea": "Autocatalytic Hysteresis Memory Cell",
  "worked": false,
  "reason": "The mechanism's quantitative predictions were not all confirmed.",
  "math_predictions": {
    "fixed_point_stability": {
      "predicted": "outer roots stable and middle root unstable",
      "observed": "derivatives [-10.2667, 5.1333, -10.2667] at gamma=-4.4741; this holds for three real roots",
      "qualification": "the roots are [-0.9324, 1.3333, 3.5990], so only two are nonnegative; the claimed nonnegative bistability is absent"
    },
    "frequency_peak": {
      "predicted": "peak near omega*tau=1",
      "observed": "maximum sampled area 12.4065 at omega*tau=0.9487",
      "status": "confirmed for this forcing"
    },
    "low_frequency_limit": {
      "predicted": "area tends to zero when omega*tau is much less than one",
      "observed": "area=106.06 at omega*tau=0.03 versus 12.41 near 0.95",
      "status": "failed; slow forcing crosses folds and produces rate-independent hysteresis"
    },
    "noise_turnover": {
      "predicted": "noise initially enhances switching then destroys coherent hysteresis",
      "observed": "area increased monotonically from 12.34 at D=0 to 13.07 at D=0.02",
      "status": "not observed in tested range"
    }
  },
  "mini_experiment": {
    "baseline_linear_accuracy": 1.0,
    "cell_accuracy": 1.0,
    "interpretation": "The toy task is trivially separable and provides no advantage for the cell."
  },
  "parameters": {"alpha": 4.0, "beta": 0.2, "seed": 1349},
  "files": ["experiment.py", "results.json", "report.json"]
}