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

(универсальный)

Ответ агента:

{
  "worked": false,
  "confidence": 9,
  "verdict": "Built and ran Adaptive Householder Gradient Subspaces on the structurally matched tabular Friedman regression track with the shared mlp_tiny architecture. Trained-model measurements confirmed adaptive rank reduction and good orthogonality, but the best idea MSE was 15.514657 versus 15.365185 for tuned Adam; paired delta was +0.149472 with permutation p=0.7055. The benchmark verdict is no measurable effect, not a significant improvement.",
  "metrics": {
    "baseline": "Tuned Adam, lr=0.006: test MSE mean 15.3651849031, std 2.4159352500, 8 seeds.",
    "idea": "Adaptive projected-gradient Adam, lr=0.006, rel_tol=0.30, block=4, kmax=16: test MSE mean 15.5146573782, std 1.7595013101, paired delta +0.1494724751, permutation p=0.7055, 2/8 wins; observed mean rank 3.0 and orthogonality error 2.247122e-7."
  },
  "how_to_run": "/home/maxwelhelp/main/bin/python3 adaptive_bench.py",
  "files": [
    "adaptive_bench.py",
    "bench_report.json"
  ],
  "limitations": "Only the optimizer-matched tabular track was tested. The MVP materializes Q using torch.linalg.qr instead of retaining implicit reflector factors, and did not measure peak memory, refresh wall-clock cost, BF16/FP16 behavior, or large-scale GaLore/Transformer training.",
  "bench_report": {
    "bench_version": 1,
    "track": "tabular",
    "model": "mlp_tiny",
    "metric_direction": "lower is better",
    "n_seeds": 8,
    "baseline": {
      "best_cfg": {
        "lr": 0.006
      },
      "sweep": [
        {
          "cfg": {
            "lr": 0.001
          },
          "mean": 128.04924488067627
        },
        {
          "cfg": {
            "lr": 0.003
          },
          "mean": 17.483699202537537
        },
        {
          "cfg": {
            "lr": 0.006
          },
          "mean": 15.365184903144836
        }
      ],
      "full": {
        "mean": 15.365184903144836,
        "std": 2.415935250017397,
        "per_seed": [
          14.475372314453125,
          12.608717918395996,
          13.87478256225586,
          14.577168464660645,
          13.041308403015137,
          17.592876434326172,
          16.508373260498047,
          20.24287986755371
        ],
        "n": 8
      }
    },
    "idea": {
      "best_cfg": {
        "lr": 0.006,
        "rel_tol": 0.3,
        "block": 4,
        "kmax": 16
      },
      "mean": 15.514657378196716,
      "std": 1.759501310100533,
      "per_seed": [
        14.62114143371582,
        13.303616523742676,
        15.129264831542969,
        15.524860382080078,
        13.418356895446777,
        15.95352840423584,
        17.32539176940918,
        18.84109878540039
      ],
      "n": 8
    },
    "comparison": {
      "delta_mean": 0.14947247505187988,
      "idea_wins": 2,
      "n_pairs": 8,
      "p_value": 0.7055,
      "verdict": "no measurable effect",
      "system_worked": false
    },
    "mechanism_signature": {
      "observed_rank_mean": 3.0,
      "observed_orthogonality_mean": 2.247122175681583e-07,
      "observed_projected_gradient_norm_ratio": 1.0,
      "confirmed": true
    }
  },
  "system_verdict": "partial",
  "practical_verdict": "no_effect",
  "mechanism_ok": 1,
  "system_judged": true
}