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

(универсальный)

Ответ агента:

{ "worked": true, "confidence": 8, "verdict": "Built Rank-Safe Variable-Projection Gauss–Newton as a local modification of the tabular mlp_tiny training procedure, using exact SVD output projection, truncated hidden Jacobian pseudoinverse, and backtracking. Against an Adam baseline tuned over the same learning-rate union, the idea achieved lower test MSE on all 8 paired seeds: mean 8.6234 versus 11.6078, paired delta -2.9844, permutation p=0.0081, so the harness verdict is idea better (significant). The neural-scale mechanism signature was not confirmed: observed full/truncated step amplification was 1.0 rather than the toy prediction, and the idea was not shown to be faster.", "metrics": { "baseline": "mlp_tiny Adam, best lr=0.01; 8-seed test MSE mean 11.607812, std 1.173520; per-seed [10.810469, 12.079232, 10.855374, 10.777785, 9.649698, 12.845276, 12.719051, 13.125610]", "idea": "Rank-safe VP-GN, best shared-grid lr=0.001; 8-seed test MSE mean 8.623417, std 1.450602; per-seed [7.800422, 11.718672, 10.022994, 6.932705, 7.960086, 8.643535, 8.414509, 7.494410]", "paired_delta": "-2.984396; idea wins 8/8 pairs; permutation p=0.0081", "sanity": "Toy singular-direction check: full step norm 1.00000008e8, truncated step norm 0.1, kept rank 1", "mechanism_signature": "Predicted amplification 1.000000064e9; trained-model observed amplification 1.0, confirmed=false" }, "bench_report": { "bench_version": 1, "track": "tabular", "model": "mlp_tiny", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.01 }, "sweep": [ { "cfg": { "lr": 0.001 }, "mean": 130.23694229125977 }, { "cfg": { "lr": 0.003 }, "mean": 17.47633147239685 }, { "cfg": { "lr": 0.01 }, "mean": 11.130715131759644 } ], "full": { "mean": 11.6078120470047, "std": 1.1735198758690657, "per_seed": [ 10.810468673706055, 12.079232215881348, 10.855374336242676, 10.777785301208496, 9.649698257446289, 12.84527587890625, 12.719051361083984, 13.1256103515625 ], "n": 8 } }, "idea": { "best_cfg": { "lr": 0.001 }, "sweep": [ { "cfg": { "lr": 0.001 }, "mean": 9.118698239326477 }, { "cfg": { "lr": 0.003 }, "mean": 10.81938362121582 }, { "cfg": { "lr": 0.01 }, "mean": 10.81938362121582 } ], "mean": 8.623416543006897, "std": 1.4506015215736154, "per_seed": [ 7.800422191619873, 11.718671798706055, 10.022994041442871, 6.932704925537109, 7.960085868835449, 8.643534660339355, 8.414508819580078, 7.494410037994385 ], "n": 8 }, "comparison": { "delta_mean": -2.9843955039978027, "idea_wins": 8, "n_pairs": 8, "per_seed_diffs": [ -3.0100464820861816, -0.36056041717529297, -0.8323802947998047, -3.8450803756713867, -1.6896123886108398, -4.2017412185668945, -4.304542541503906, -5.631200313568115 ], "p_value": 0.0081, "mde": 1.5535327571979936, "mde_rel_pct": 13.383510612569488, "verdict": "idea better (significant)", "system_worked": true }, "mechanism_signature": { "predicted_full_to_truncated_amplification": 1000000064.0, "observed_full_step_norm": 0.04894990101456642, "observed_truncated_step_norm": 0.04894990101456642, "observed_amplification": 1.0, "confirmed": false } }, "how_to_run": "/home/maxwelhelp/main/bin/python3 bench_vp_gn.py", "files": [ "bench_vp_gn.py", "bench_report.json", "bench_stdout.txt" ], "limitations": "Only the structurally matched tabular/Friedman regression track was tested. The dense hidden Jacobian was limited to the tiny MLP and 400-sample subsets; wall-clock speed was not the primary comparison, no vision/sequence/dynamics tracks were run, and the trained-model signature did not reproduce the toy singular-direction amplification.", "system_verdict": "worked", "practical_verdict": "helps", "mechanism_ok": 0, "system_judged": true }