# Эксперимент: Finite-Candidate Neural Reference Shield (#1076)

{
  "worked": true,
  "confidence": 9,
  "verdict": "Built a deterministic finite-candidate KKT shield for a 2D actuator set defined by disk and box constraints, with projected repair and SLSQP references. The mechanism manifested: interior commands were preserved exactly, the measured transition was 1.01 versus the predicted 1.00, radius-transition slope was 1.00 versus predicted 1.00, maximum violation was 6.7e-16, and maximum error versus SLSQP was 2.7e-8. The shield was faster than iterative repair in this toy test, at 0.72 ms versus 6.57 ms per command, with slightly smaller command deviation.",
  "metrics": {
    "baseline": "Projected repair: mean violation 1.67e-17, mean distance 0.6012, 6.57 ms/command; raw mean violation 1.6688.",
    "idea": "Finite-candidate shield: mean violation 5.16e-17, mean distance 0.5934, 0.72 ms/command; interior maximum error 0, transition 1.01 versus 1.00 predicted, radius slope 1.00 versus 1.00 predicted, maximum SLSQP error 2.71e-8."
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 shield_experiment.py",
  "files": [
    "shield_experiment.py",
    "results.json"
  ],
  "limitations": "Only a convex 2D disk-and-box actuator region was tested. General nonconvex nonlinear constraints, higher-dimensional active-set enumeration, neural-policy training, state noise, model mismatch, and GPU execution were not tested."
}