{ "bench_version": 1, "track": "tabular", "model": "custom_branch_mlp", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.01, "weight_decay": 0.0 }, "sweep": [ { "cfg": { "lr": 0.001, "weight_decay": 0.0 }, "mean": 12.950589418411255 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 6.888427972793579 }, { "cfg": { "lr": 0.01, "weight_decay": 0.0 }, "mean": 6.682573318481445 } ], "full": { "mean": 6.839571237564087, "std": 0.3468641081837492, "per_seed": [ 6.566773414611816, 6.700428485870361, 6.708205223083496, 6.754886150360107, 6.259535312652588, 7.132112979888916, 7.337502479553223, 7.2571258544921875 ], "n": 8 } }, "idea": { "best_cfg": { "lr": 0.01, "weight_decay": 0.0 }, "sweep": [ { "cfg": { "lr": 0.001, "weight_decay": 0.0 }, "mean": 13.624429821968079 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 7.040245294570923 }, { "cfg": { "lr": 0.01, "weight_decay": 0.0 }, "mean": 6.819535434246063 } ], "per_seed": [ 6.49015998840332, 6.686732769012451, 6.5792107582092285, 6.80715799331665, 6.233836650848389, 7.175884246826172, 7.312864780426025, 7.2704362869262695 ], "mean": 6.819535434246063, "std": 0.3709357840355057, "n": 8 }, "comparison": { "delta_mean": -0.02003580331802368, "idea_wins": 5, "n_pairs": 8, "per_seed_diffs": [ -0.0766134262084961, -0.013695716857910156, -0.12899446487426758, 0.05227184295654297, -0.02569866180419922, 0.04377126693725586, -0.024637699127197266, 0.013310432434082031 ], "p_value": 0.4109, "mde": 0.05054426158733883, "mde_rel_pct": 0.7389975165364338, "verdict": "no measurable effect", "system_worked": false }, "mechanism_signature": { "math_check": { "epsilon_values": [ 0.001, 0.002130941015366798, 0.004540909610972475, 0.009676410537094535, 0.020619860095022202, 0.043939705607607904, 0.09363292088239417, 0.19952623149688797 ], "remainder_values": [ 1.9010061570122444e-08, 1.8393117422730257e-07, 1.7794226864628757e-06, 1.7210834407346843e-05, 0.00016639443047186702, 0.001607205940536596, 0.01551344151929597, 0.15093468018953293 ], "observed_remainder_slope": 2.9993339395602745, "predicted_remainder_slope": 3.0, "gram_invariance_max_error": 5.1514348342607263e-14, "prediction_pass": true }, "trained_model": { "source": "trained baseline model on tabular test inputs", "epsilon_values": [ 0.0031622776601683794, 0.006072021956909884, 0.011659144011798317, 0.0223872113856834, 0.042986623470822764, 0.08254041852680181, 0.15848931924611134 ], "observed_remainder_slope": 3.0196060096593453, "predicted_remainder_slope": 3.0, "permutation_max_error": 1.9073486328125e-06, "confirmed": true }, "custom_track": null }, "protocol_notes": { "paired_seeds": [ 0, 1, 2, 3, 4, 5, 6, 7 ], "n_train": 800, "n_test": 400, "architecture": "same width-64 exchangeable branch MLP; only phi vs quadratic Taylor block differs", "baseline_grid": [ { "lr": 0.001, "weight_decay": 0.0 }, { "lr": 0.003, "weight_decay": 0.0 }, { "lr": 0.01, "weight_decay": 0.0 } ], "idea_grid": [ { "lr": 0.001, "weight_decay": 0.0 }, { "lr": 0.003, "weight_decay": 0.0 }, { "lr": 0.01, "weight_decay": 0.0 } ] } }