# Стенд-проверка (stage-2) · промт оператора:

(универсальный)

Ответ агента:

{
  "worked": false,
  "confidence": 10,
  "verdict": "The feasibility-preserving compensator was implemented as a matched rnn_small system on the registered dynamics track. Its mechanism signature was confirmed: predicted spectral decay 0.88227 versus observed 0.88158, with anti-windup peak magnitude 0.16032 versus ordinary integral 0.70819. However, it significantly lost on rollout MSE, so the idea did not transfer successfully to neural benchmark training.",
  "metrics": {
    "baseline": "mean MSE 0.0014198302378645167; best lr 0.01, epochs 12",
    "idea": "mean MSE 0.47640712931752205; best KI 0.05, KA 0.7, lr 0.01, epochs 12",
    "comparison": "delta_mean +0.47498729907965753; permutation p=0.0081; idea wins 0/8"
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 bench_experiment.py",
  "files": [
    "bench_experiment.py",
    "bench_report.json",
    "math_check.json"
  ],
  "limitations": "Only the registered dynamics track and rnn_small model were tested; no noisy measurements, multidimensional task space, explicit actuator constraint suite, or longer-horizon deployment was evaluated.",
  "bench_report": {
    "bench_version": 1,
    "track": "dynamics",
    "model": "rnn_small",
    "metric_direction": "lower is better",
    "n_seeds": 8,
    "baseline": {
      "best_cfg": {
        "lr": 0.01,
        "epochs": 12
      },
      "sweep": [
        {
          "cfg": {
            "lr": 0.001,
            "epochs": 12
          },
          "mean": 0.005020207085181028
        },
        {
          "cfg": {
            "lr": 0.003,
            "epochs": 12
          },
          "mean": 0.0029489120497601107
        },
        {
          "cfg": {
            "lr": 0.01,
            "epochs": 12
          },
          "mean": 0.0014198302378645167
        }
      ],
      "full": {
        "mean": 0.0014198302378645167,
        "std": 0.00039839114901047844,
        "per_seed": [
          0.0009327387670055032,
          0.0018641706556081772,
          0.0010785290505737066,
          0.0020708823576569557,
          0.0009260809747502208,
          0.0014055331703275442,
          0.0016443432541564107,
          0.001436363672837615
        ],
        "n": 8
      }
    },
    "idea": {
      "best_cfg": {
        "lr": 0.01,
        "epochs": 12,
        "ki": 0.05,
        "ka": 0.7
      },
      "mean": 0.47640712931752205,
      "std": 0.04352936722428276,
      "per_seed": [
        0.4626535177230835,
        0.45927271246910095,
        0.5099033713340759,
        0.5794274210929871,
        0.4526548683643341,
        0.45540836453437805,
        0.4523857831954956,
        0.4395509958267212
      ],
      "n": 8
    },
    "comparison": {
      "delta_mean": 0.47498729907965753,
      "idea_wins": 0,
      "n_pairs": 8,
      "per_seed_diffs": [
        0.461720778956078,
        0.4574085418134928,
        0.5088248422835022,
        0.5773565387353301,
        0.4517287873895839,
        0.4540028313640505,
        0.4507414399413392,
        0.4381146321538836
      ],
      "p_value": 0.0081,
      "verdict": "idea worse (significant)",
      "system_worked": false
    },
    "mechanism_signature": {
      "math_and_trained_behavior": {
        "spectral_radius_predicted": 0.8822698000045112,
        "decay_factor_observed": 0.8815778323329195,
        "decay_relative_error": 0.00078430393014483,
        "ordinary_max_compensator": 0.708185009625288,
        "antiwindup_max_compensator": 0.1603214553812824,
        "windup_reduction_ratio": 0.22638357661102013,
        "constraint_violation": 0.0,
        "prediction_confirmed": true,
        "trained_nominal_abs_mean": 0.6836502552032471,
        "trained_shaped_abs_mean": 0.21633578836917877,
        "trained_residual_abs_mean": 0.2636767625808716,
        "trained_compensator_abs_mean": 0.22096218168735504,
        "confirmed": true
      },
      "idea_sweep": [
        {
          "cfg": {
            "lr": 0.01,
            "epochs": 12,
            "ki": 0.05,
            "ka": 0.7
          },
          "mean": 0.47640712931752205,
          "std": 0.04352936722428276,
          "n": 8
        },
        {
          "cfg": {
            "lr": 0.01,
            "epochs": 12,
            "ki": 0.075,
            "ka": 0.9
          },
          "mean": 0.4816158339381218,
          "std": 0.04396653306668558,
          "n": 8
        },
        {
          "cfg": {
            "lr": 0.01,
            "epochs": 12,
            "ki": 0.1,
            "ka": 1.1
          },
          "mean": 0.48671987280249596,
          "std": 0.044320348769445794,
          "n": 8
        }
      ]
    }
  },
  "system_verdict": "failed",
  "practical_verdict": "harms",
  "mechanism_ok": 0,
  "system_judged": true
}