{ "bench_version": 1, "track": "tabular", "model": "mlp_tiny", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.006, "weight_decay": 0.0 }, "sweep": [ { "cfg": { "lr": 0.0015, "weight_decay": 0.0 }, "mean": 20.46571445465088 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 16.009180068969727 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0 }, "mean": 13.129557490348816 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0001 }, "mean": 15.930055141448975 } ], "full": { "mean": 13.129557490348816, "std": 1.5250596544234398, "per_seed": [ 13.93191909790039, 13.616814613342285, 10.975268363952637, 13.225879669189453, 10.728046417236328, 12.577923774719238, 14.791933059692383, 15.188674926757812 ], "n": 8 } }, "idea": { "best_cfg": { "lr": 0.006, "weight_decay": 0.0 }, "sweep": [ { "cfg": { "lr": 0.0015, "weight_decay": 0.0 }, "mean": 19.026683568954468, "std": 2.310262868739491, "per_seed": [ 20.679960250854492, 15.556901931762695, 18.93318748474121, 17.145143508911133, 19.445880889892578, 16.939167022705078, 20.18524742126465, 23.327980041503906 ], "signature_per_seed": [ { "accepted": 35, "rejected": 25, "predicted_decrease_mean": 1.382528269290924, "observed_decrease_mean": 1.3716121673583985, "radius_final": 0.012100000000000001 }, { "accepted": 41, "rejected": 19, "predicted_decrease_mean": 2.0615332511866966, "observed_decrease_mean": 2.022140456408989, "radius_final": 0.014641000000000003 }, { "accepted": 40, "rejected": 20, "predicted_decrease_mean": 1.1156054258346557, "observed_decrease_mean": 1.1034204244613648, "radius_final": 0.011000000000000001 }, { "accepted": 41, "rejected": 19, "predicted_decrease_mean": 1.26738692674695, "observed_decrease_mean": 1.255918735411109, "radius_final": 0.011000000000000001 }, { "accepted": 38, "rejected": 22, "predicted_decrease_mean": 1.2735088334271782, "observed_decrease_mean": 1.2580657758210834, "radius_final": 0.01 }, { "accepted": 40, "rejected": 20, "predicted_decrease_mean": 1.3169351086020469, "observed_decrease_mean": 1.3040775299072265, "radius_final": 0.01 }, { "accepted": 41, "rejected": 19, "predicted_decrease_mean": 1.2663456913901538, "observed_decrease_mean": 1.2509966594416921, "radius_final": 0.013310000000000002 }, { "accepted": 45, "rejected": 15, "predicted_decrease_mean": 1.2946510824892257, "observed_decrease_mean": 1.281169213189019, "radius_final": 0.01806992220000001 } ] }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 15.578829765319824, "std": 1.2037474931922048, "per_seed": [ 16.088903427124023, 13.419285774230957, 15.963817596435547, 14.6215238571167, 14.568916320800781, 15.911458969116211, 17.396286010742188, 16.660446166992188 ], "signature_per_seed": [ { "accepted": 32, "rejected": 28, "predicted_decrease_mean": 1.2606446049176157, "observed_decrease_mean": 1.1952399015426636, "radius_final": 0.0102 }, { "accepted": 33, "rejected": 27, "predicted_decrease_mean": 1.7287133065136997, "observed_decrease_mean": 1.6648587313565342, "radius_final": 0.012342000000000002 }, { "accepted": 38, "rejected": 22, "predicted_decrease_mean": 1.2891257283718962, "observed_decrease_mean": 1.2490642447220652, "radius_final": 0.011000000000000001 }, { "accepted": 45, "rejected": 15, "predicted_decrease_mean": 1.1980243287152714, "observed_decrease_mean": 1.1577167299058702, "radius_final": 0.011000000000000001 }, { "accepted": 35, "rejected": 25, "predicted_decrease_mean": 1.508543290410723, "observed_decrease_mean": 1.4751929146902902, "radius_final": 0.01 }, { "accepted": 35, "rejected": 25, "predicted_decrease_mean": 1.1945056153195246, "observed_decrease_mean": 1.1530200685773577, "radius_final": 0.01 }, { "accepted": 36, "rejected": 24, "predicted_decrease_mean": 1.228395100062092, "observed_decrease_mean": 1.1819042099846735, "radius_final": 0.01 }, { "accepted": 43, "rejected": 17, "predicted_decrease_mean": 1.962806150772024, "observed_decrease_mean": 1.9334695505541424, "radius_final": 0.014933820000000006 } ] }, { "cfg": { "lr": 0.006, "weight_decay": 0.0 }, "mean": 12.88543701171875, "std": 1.4828934325502452, "per_seed": [ 13.835183143615723, 11.309344291687012, 12.560839653015137, 12.500975608825684, 10.138545989990234, 13.836167335510254, 14.998884201049805, 13.903555870056152 ], "signature_per_seed": [ { "accepted": 28, "rejected": 32, "predicted_decrease_mean": 1.379367763974837, "observed_decrease_mean": 1.2420953001294817, "radius_final": 0.010404000000000002 }, { "accepted": 26, "rejected": 34, "predicted_decrease_mean": 1.344854632822367, "observed_decrease_mean": 1.24173248731173, "radius_final": 0.01 }, { "accepted": 33, "rejected": 27, "predicted_decrease_mean": 1.2900403211965705, "observed_decrease_mean": 1.2163373918244333, "radius_final": 0.011000000000000001 }, { "accepted": 36, "rejected": 24, "predicted_decrease_mean": 1.089351099398401, "observed_decrease_mean": 1.0336363050672743, "radius_final": 0.011000000000000001 }, { "accepted": 38, "rejected": 22, "predicted_decrease_mean": 2.0978008941502164, "observed_decrease_mean": 1.8838776789213483, "radius_final": 0.01 }, { "accepted": 30, "rejected": 30, "predicted_decrease_mean": 1.233123182753722, "observed_decrease_mean": 1.1296147982279459, "radius_final": 0.01 }, { "accepted": 33, "rejected": 27, "predicted_decrease_mean": 1.392147899577112, "observed_decrease_mean": 1.2645280722415808, "radius_final": 0.01 }, { "accepted": 39, "rejected": 21, "predicted_decrease_mean": 1.6642315091613011, "observed_decrease_mean": 1.6168656226916192, "radius_final": 0.011220000000000003 } ] }, { "cfg": { "lr": 0.003, "weight_decay": 0.0001 }, "mean": 15.578837275505066, "std": 1.203698466172168, "per_seed": [ 16.08891487121582, 13.419283866882324, 15.963841438293457, 14.621552467346191, 14.569123268127441, 15.911477088928223, 17.396265029907227, 16.660240173339844 ], "signature_per_seed": [ { "accepted": 32, "rejected": 28, "predicted_decrease_mean": 1.2606316241435707, "observed_decrease_mean": 1.1952308416366577, "radius_final": 0.0102 }, { "accepted": 33, "rejected": 27, "predicted_decrease_mean": 1.7287147830833087, "observed_decrease_mean": 1.6648634708288945, "radius_final": 0.012342000000000002 }, { "accepted": 38, "rejected": 22, "predicted_decrease_mean": 1.2891209054934352, "observed_decrease_mean": 1.2490592253835577, "radius_final": 0.011000000000000001 }, { "accepted": 45, "rejected": 15, "predicted_decrease_mean": 1.1980234457386865, "observed_decrease_mean": 1.1577159881591796, "radius_final": 0.011000000000000001 }, { "accepted": 35, "rejected": 25, "predicted_decrease_mean": 1.5085362383297511, "observed_decrease_mean": 1.475187737601144, "radius_final": 0.01 }, { "accepted": 35, "rejected": 25, "predicted_decrease_mean": 1.1944936403206416, "observed_decrease_mean": 1.1530044828142438, "radius_final": 0.01 }, { "accepted": 36, "rejected": 24, "predicted_decrease_mean": 1.228386132667462, "observed_decrease_mean": 1.1818928718566895, "radius_final": 0.01 }, { "accepted": 43, "rejected": 17, "predicted_decrease_mean": 1.9628009385277712, "observed_decrease_mean": 1.933466112890909, "radius_final": 0.014933820000000006 } ] } ], "mean": 12.88543701171875, "std": 1.4828934325502452, "per_seed": [ 13.835183143615723, 11.309344291687012, 12.560839653015137, 12.500975608825684, 10.138545989990234, 13.836167335510254, 14.998884201049805, 13.903555870056152 ], "n": 8, "full": { "mean": 12.88543701171875, "std": 1.4828934325502452, "per_seed": [ 13.835183143615723, 11.309344291687012, 12.560839653015137, 12.500975608825684, 10.138545989990234, 13.836167335510254, 14.998884201049805, 13.903555870056152 ], "n": 8 } }, "comparison": { "delta_mean": -0.24412047863006592, "idea_wins": 5, "n_pairs": 8, "per_seed_diffs": [ -0.09673595428466797, -2.3074703216552734, 1.5855712890625, -0.7249040603637695, -0.5895004272460938, 1.2582435607910156, 0.20695114135742188, -1.2851190567016602 ], "p_value": 0.6102, "mde": 1.0727371153900689, "mde_rel_pct": 8.170398097411958, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "mechanism_signature": { "predicted_decrease_mean": 1.4363646628793159, "observed_decrease_mean": 1.3285859570519265, "observed_to_predicted_ratio": 0.9261847601497202, "accepted_steps_mean": 32.875, "rejected_steps_mean": 27.125, "confirmed": true }, "protocol": { "paired_seeds": [ 0, 1, 2, 3, 4, 5, 6, 7 ], "epochs": 15, "train_samples": 400, "test_samples": 200, "grid": [ { "lr": 0.0015, "weight_decay": 0.0 }, { "lr": 0.003, "weight_decay": 0.0 }, { "lr": 0.006, "weight_decay": 0.0 }, { "lr": 0.003, "weight_decay": 0.0001 } ] }, "runtime_sec": 35.358980655670166 } }