{ "bench_version": 1, "track": "tabular", "model": "mlp_tiny", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.01, "weight_decay": 0.0001 }, "sweep": [ { "cfg": { "lr": 0.001, "weight_decay": 0.0 }, "mean": 130.23694229125977 }, { "cfg": { "lr": 0.001, "weight_decay": 0.0001 }, "mean": 130.24530410766602 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 17.47633147239685 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0001 }, "mean": 17.535193920135498 }, { "cfg": { "lr": 0.01, "weight_decay": 0.0 }, "mean": 11.130715131759644 }, { "cfg": { "lr": 0.01, "weight_decay": 0.0001 }, "mean": 11.037371158599854 } ], "full": { "mean": 11.380033016204834, "std": 1.038192373606595, "per_seed": [ 10.492472648620605, 12.06095027923584, 10.988505363464355, 10.607556343078613, 9.646132469177246, 11.897088050842285, 12.67528247833252, 12.672276496887207 ], "n": 8 } }, "idea": { "mean": 19.045888423919678, "std": 11.022546935963494, "per_seed": [ 10.960366249084473, 18.753145217895508, 37.41089630126953, 12.765547752380371, 9.869215965270996, 14.048157691955566, 10.755529403686523, 37.80424880981445 ], "n": 8 }, "comparison": { "delta_mean": 7.665855407714844, "idea_wins": 1, "n_pairs": 8, "per_seed_diffs": [ 0.4678936004638672, 6.692194938659668, 26.422390937805176, 2.157991409301758, 0.22308349609375, 2.1510696411132812, -1.919753074645996, 25.131972312927246 ], "p_value": 0.0394, "mde": 9.573659179917387, "mde_rel_pct": 84.12681374724288, "verdict": "idea worse (significant)", "system_worked": false }, "mechanism_signature": { "prediction": "skew component is orthogonal to current gradient and increases complex/rotational response while stability requires rho<1", "observed_jacobian_rayleigh": -1.708630919456482, "observed_euler_rayleigh_factor": 1.0051258927583695, "observed_rotation_gradient_cosine": -5.9046900702242056e-09, "observed_skew_to_gradient_norm": 0.5999951958656311, "confirmed": true }, "idea_sweep": [ { "cfg": { "lr": 0.01, "alpha": 0.0 }, "mean": 29.906624674797058, "std": 26.379266590239755, "per_seed": [ 39.70406723022461, 17.099863052368164, 10.755126953125, 14.188899040222168, 11.864416122436523, 10.44906234741211, 43.637142181396484, 91.5544204711914 ] }, { "cfg": { "lr": 0.01, "alpha": 0.3 }, "mean": 21.96018660068512, "std": 21.920466510288684, "per_seed": [ 11.32851791381836, 15.652362823486328, 12.332928657531738, 13.191065788269043, 18.957841873168945, 13.081674575805664, 11.51638126373291, 79.62071990966797 ] }, { "cfg": { "lr": 0.01, "alpha": 0.6 }, "mean": 19.045888423919678, "std": 11.022546935963494, "per_seed": [ 10.960366249084473, 18.753145217895508, 37.41089630126953, 12.765547752380371, 9.869215965270996, 14.048157691955566, 10.755529403686523, 37.80424880981445 ] } ], "budget": { "epochs": 12, "batch": 128, "n_train": 400, "n_test": 400, "seconds": 17.008500576019287 }, "custom_track": null }