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

(универсальный)

Ответ агента:

{
  "worked": false,
  "confidence": 9,
  "verdict": "Implemented a phase-space kinetic mixing layer and evaluated it against a raw-token GRU using the canonical trainer, parity-matched learning-rate sweep, and eight paired seeds. The kinetic model was significantly worse: test MSE 0.00132353 versus 0.00102014 for baseline, paired delta +0.00030339, permutation p=0.04965. The diffusion attenuation check was close, but the transport displacement measurement was confounded by learned gating and channel mixing; regardless, the trained benchmark metric is a clear loss.",
  "metrics": {
    "baseline": "dynamics raw-token GRU; best lr=0.006; 8-seed test MSE mean=0.00102014, std=0.00017824",
    "idea": "dynamics kinetic GRU; h=0.08, best lr=0.006; 8-seed test MSE mean=0.00132353, std=0.00057783",
    "paired_delta": "0.0003033862 (idea minus baseline; positive is worse)",
    "permutation_p_value": 0.04965,
    "idea_wins": 1,
    "n_pairs": 8,
    "parameter_counts": "baseline 3585, idea 3609"
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 stage2_bench.py",
  "files": [
    "stage2_bench.py",
    "bench_report.json",
    "stage2_output.txt"
  ],
  "limitations": "Only the built-in dynamics track was tested; long-horizon rollout error, irregular-coordinate diffusion, multiscale stacks, and additional tracks were not tested. The transport mechanism signature was confounded by the learned gate and channel mixing and should not be treated as a clean independent confirmation.",
  "bench_report": {
    "bench_version": 1,
    "track": "dynamics",
    "model": "rnn_small_kinetic_vs_raw",
    "metric_direction": "lower is better",
    "n_seeds": 8,
    "baseline": {
      "best_cfg": {
        "lr": 0.006,
        "h": null
      },
      "sweep": [
        {
          "cfg": {
            "lr": 0.0015,
            "h": null
          },
          "mean": 0.0033734575
        },
        {
          "cfg": {
            "lr": 0.003,
            "h": null
          },
          "mean": 0.0019945995
        },
        {
          "cfg": {
            "lr": 0.006,
            "h": null
          },
          "mean": 0.0011413919
        }
      ],
      "full": {
        "mean": 0.00102014,
        "std": 0.0001782361,
        "per_seed": [
          0.0009858898,
          0.0012092517,
          0.0009726743,
          0.0013977519,
          0.0009046853,
          0.0009852344,
          0.0008830014,
          0.0008226314
        ],
        "n": 8
      },
      "union_lr_grid": [
        0.0015,
        0.003,
        0.006
      ]
    },
    "idea": {
      "mean": 0.0013235263,
      "std": 0.0005778316,
      "per_seed": [
        0.0010875554,
        0.0014675898,
        0.0012377128,
        0.0027645107,
        0.0009836847,
        0.0007464993,
        0.0011154669,
        0.0011851905
      ],
      "n": 8
    },
    "comparison": {
      "delta_mean": 0.0003033862,
      "idea_wins": 1,
      "n_pairs": 8,
      "p_value": 0.04965,
      "verdict": "idea worse (significant)",
      "system_worked": false
    },
    "mechanism_signature": {
      "prediction": "x Fourier attenuation approximately exp(-h*k^2) and transport displacement proportional to h*x",
      "predicted_fourier_attenuation": 0.4867523,
      "observed_fourier_attenuation": 0.470259,
      "attenuation_relative_error": 0.0338844,
      "predicted_transport_displacement": 1.28,
      "observed_transport_displacement": -0.1354923,
      "confirmed": true
    },
    "idea_sweep": [
      {
        "cfg": {
          "lr": 0.0015,
          "h": 0.08
        },
        "mean": 0.0084915037
      },
      {
        "cfg": {
          "lr": 0.003,
          "h": 0.08
        },
        "mean": 0.0029999098
      },
      {
        "cfg": {
          "lr": 0.006,
          "h": 0.08
        },
        "mean": 0.0016393422
      }
    ],
    "best_idea_cfg": {
      "lr": 0.006,
      "h": 0.08
    },
    "parameter_parity": {
      "baseline": 3585,
      "idea": 3609
    },
    "custom_track": null
  },
  "system_verdict": "partial",
  "practical_verdict": "harms",
  "mechanism_ok": 1,
  "system_judged": true
}