# Эксперимент: Spectral-Pole-Tuned Decentralized Optimizer (#691)

{
  "worked": false,
  "confidence": 10,
  "verdict": "Implemented the exact modal dynamics, spectral gain formulas, stability sweeps, and a small quadratic decentralized-tracking comparison in spectral_pole_experiment.py. The paper-consistent convention κ=λ₂/λ_N reproduced the predicted worst pole radius to numerical precision and showed endpoint equalization, but the authoritative idea defines κ=λ_N/λ₂ while retaining the same formula; under that literal definition, measured radii were 2.63, 9.13, 20.09, 63.11, and 209.48 for condition ratios 2, 5, 10, 30, and 100, respectively, causing divergence. Therefore the proposed idea does not work as stated, despite the corrected paper convention exhibiting the claimed mechanism.",
  "metrics": {
    "baseline": "Quadratic toy baseline (alpha=0.1, epsilon=0.1): final loss 0.2039, finite after 100 steps.",
    "idea": "Literal spectral gains on an 8-node path: final loss exceeded 1.21e101 and the modal worst radius was 9.13 for graph condition ratio 25.27; paper-consistent gains gave radius 0.9711 for ratio 25.27 and final loss 0.2395."
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 spectral_pole_experiment.py",
  "files": [
    "spectral_pole_experiment.py",
    "results.json"
  ],
  "limitations": "No neural-network, MNIST/CIFAR, communication-byte, GPU, or gain-grid experiment was run because the literal formula failed the required toy mechanism check. The implementation uses exact eigenspectrum values in the toy graph rather than noisy distributed eigenvalue estimation."
}