# Эксперимент: ISS-CLF/RCBF Neural Policy Shield (#596)

{
  "worked": true,
  "confidence": 8,
  "verdict": "Built a scalar robust ISS-CLF/CBF policy shield with box-bounded control, analytic disturbance margins, slacked QP projection, and deterministic sweeps. The mechanism manifested: robust CLF margin scaled linearly with disturbance at observed slope 0.600 versus predicted 0.600, zero-slack feasibility ended in the predicted interval [0.80, 0.85] around the analytic boundary 0.80, and the adversarial trajectory stayed within the exponential Lyapunov envelope with max ratio 0.9999998 and residual 1.22e-6. The shield reduced safety violation fraction from 0.917 to 0 while requiring substantial action deviation, so this is evidence for safety behavior but not general RL performance.",
  "metrics": {
    "baseline": "Adversarial scalar episode: violation_fraction=0.917, final_abs_x=3.2000.",
    "idea": "Scaling slope=0.600 vs predicted 0.600; feasibility boundary predicted wbar=0.800 and observed grid transition between 0.800 and 0.850; envelope max V/bound=0.999999845, max robust CLF residual=1.22e-6; shield episode violation_fraction=0.000, final_abs_x=0.5362, mean action deviation=1.3319, mean slack=1.02e-6."
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 shield_experiment.py",
  "files": [
    "shield_experiment.py",
    "results.json"
  ],
  "limitations": "Only a scalar known plant was tested; no neural policy training, differentiable QP backpropagation, learned dynamics, multidimensional obstacles, stochastic seeds, runtime profiling, or PPO/SAC comparison was performed. The QP uses active-set enumeration rather than a production solver, and the feasibility sweep resolution is 0.05."
}