{ "bench_version": 1, "track": "vision", "model": "cnn_small", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.003, "wd": 0.0 }, "sweep": [ { "cfg": { "lr": 0.001, "wd": 0.0 }, "mean": 0.7899999767541885 }, { "cfg": { "lr": 0.001, "wd": 0.0001 }, "mean": 0.7899999767541885 }, { "cfg": { "lr": 0.003, "wd": 0.0 }, "mean": 0.7824999839067459 }, { "cfg": { "lr": 0.003, "wd": 0.0001 }, "mean": 0.8112499862909317 }, { "cfg": { "lr": 0.01, "wd": 0.0 }, "mean": 0.8837499767541885 }, { "cfg": { "lr": 0.01, "wd": 0.0001 }, "mean": 0.8912499696016312 } ], "full": { "mean": 0.7974999845027924, "std": 0.031622772360742395, "per_seed": [ 0.7799999713897705, 0.7599999904632568, 0.8399999737739563, 0.75, 0.8149999976158142, 0.7899999618530273, 0.8399999737739563, 0.8050000071525574 ], "n": 8 } }, "idea": { "mean": 0.778749980032444, "std": 0.04196351799020234, "per_seed": [ 0.7549999952316284, 0.6850000023841858, 0.7949999570846558, 0.7799999713897705, 0.8149999976158142, 0.7699999809265137, 0.7999999523162842, 0.8299999833106995 ], "n": 8 }, "comparison": { "delta_mean": -0.018750004470348358, "idea_wins": 5, "n_pairs": 8, "per_seed_diffs": [ -0.02499997615814209, -0.07499998807907104, -0.04500001668930054, 0.029999971389770508, 0.0, -0.019999980926513672, -0.04000002145767212, 0.02499997615814209 ], "p_value": 0.20525, "mde": 0.029961011266907856, "mde_rel_pct": 3.7568666895444873, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "best_cfg": { "lr": 0.003, "wd": 0.0, "p": 0.5 }, "p": 0.5, "q_leaf_observed": [ 0.6304718255996704, 0.596589982509613, 0.7500549554824829, 0.7355444431304932, 0.4830438196659088, 0.7880200147628784, 0.60601806640625, 0.7255614995956421 ], "predicted_zero": [ 0.6304718255996704, 0.596589982509613, 0.7500549554824829, 0.7355444431304932, 0.4830438196659088, 0.7880200147628784, 0.60601806640625, 0.7255614995956421 ], "observed_zero": [ 0.8817968964576721, 0.6903906464576721, 0.816162109375, 0.8238997459411621, 0.8374186158180237, 0.7389615774154663, 0.8830631375312805, 0.8237109184265137 ], "confirmed": false, "mean_abs_prediction_error": 0.1597769893705845 }, "idea_sweep": [ { "cfg": { "lr": 0.003, "wd": 0.0, "p": 0.25 }, "mean": 0.7931249737739563, "per_seed": [ 0.7749999761581421, 0.7899999618530273, 0.8050000071525574, 0.7699999809265137, 0.7949999570846558, 0.8449999690055847, 0.7999999523162842, 0.7649999856948853 ] }, { "cfg": { "lr": 0.003, "wd": 0.0, "p": 0.5 }, "mean": 0.778749980032444, "per_seed": [ 0.7549999952316284, 0.6850000023841858, 0.7949999570846558, 0.7799999713897705, 0.8149999976158142, 0.7699999809265137, 0.7999999523162842, 0.8299999833106995 ] }, { "cfg": { "lr": 0.003, "wd": 0.0, "p": 0.75 }, "mean": 0.7874999865889549, "per_seed": [ 0.7749999761581421, 0.8700000047683716, 0.7699999809265137, 0.75, 0.8349999785423279, 0.7649999856948853, 0.8050000071525574, 0.7299999594688416 ] } ], "structural_match": "vision CNN spatial hierarchical pooling", "epochs": 8, "n_train": 400, "n_test": 200 }