{ "bench_version": 1, "track": "tabular", "model": "mlp_tiny", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.006, "beta": 0.9 }, "sweep": [ { "cfg": { "lr": 0.001, "beta": 0.8 }, "mean": 137.9729061126709 }, { "cfg": { "lr": 0.001, "beta": 0.9 }, "mean": 138.0125026702881 }, { "cfg": { "lr": 0.003, "beta": 0.8 }, "mean": 45.94260215759277 }, { "cfg": { "lr": 0.003, "beta": 0.9 }, "mean": 45.709683418273926 }, { "cfg": { "lr": 0.006, "beta": 0.8 }, "mean": 20.55884027481079 }, { "cfg": { "lr": 0.006, "beta": 0.9 }, "mean": 20.514626026153564 } ], "full": { "mean": 21.575043439865112, "std": 1.9157591398399918, "per_seed": [ 22.02391815185547, 17.909542083740234, 21.564067840576172, 20.560976028442383, 20.68805694580078, 23.367746353149414, 21.641395568847656, 24.84464454650879 ], "n": 8 } }, "idea": { "mean": 21.585941553115845, "std": 1.8841984845237936, "per_seed": [ 22.23833465576172, 17.9639835357666, 21.634672164916992, 20.441940307617188, 20.853069305419922, 23.09563446044922, 21.598304748535156, 24.86159324645996 ], "n": 8 }, "comparison": { "delta_mean": 0.010898113250732422, "idea_wins": 3, "n_pairs": 8, "per_seed_diffs": [ 0.21441650390625, 0.05444145202636719, 0.07060432434082031, -0.11903572082519531, 0.16501235961914062, -0.2721118927001953, -0.0430908203125, 0.016948699951171875 ], "p_value": 0.8426, "mde": 0.13051985349480585, "mde_rel_pct": 0.6049575467071313, "verdict": "no measurable effect", "system_worked": false }, "mechanism_signature": { "prediction": "heldout alignment gating routes post when rho>=tau and reduces negative heldout alignment", "observed": { "post_fraction": 0.9606481481481481, "rho": 0.2859362520431751, "post_negative": 0.039351851851851846, "chosen_negative": 0.03306878306878307, "predicted_fraction": 0.9606481481481481, "confirmed": true }, "tau": 0.0, "trained_models": true }, "idea_sweep": [ { "cfg": { "lr": 0.001, "beta": 0.9, "tau": 0.0 }, "mean": 136.1921682357788, "per_seed": [ 149.82797241210938, 138.97898864746094, 137.3268585205078, 125.91619110107422, 135.9979705810547, 124.63673400878906, 128.59912109375, 148.25350952148438 ] }, { "cfg": { "lr": 0.001, "beta": 0.9, "tau": 0.05 }, "mean": 136.1921682357788, "per_seed": [ 149.82797241210938, 138.97898864746094, 137.3268585205078, 125.91619110107422, 135.9979705810547, 124.63673400878906, 128.59912109375, 148.25350952148438 ] }, { "cfg": { "lr": 0.001, "beta": 0.9, "tau": 0.1 }, "mean": 136.7702980041504, "per_seed": [ 150.35202026367188, 139.48507690429688, 137.9378662109375, 125.91619110107422, 136.5552520751953, 125.13619232177734, 129.91458129882812, 148.86520385742188 ] }, { "cfg": { "lr": 0.003, "beta": 0.9, "tau": 0.0 }, "mean": 44.54532766342163, "per_seed": [ 53.77455520629883, 45.56505584716797, 42.8125, 40.686622619628906, 39.3796501159668, 41.54080581665039, 41.00275421142578, 51.600677490234375 ] }, { "cfg": { "lr": 0.003, "beta": 0.9, "tau": 0.05 }, "mean": 44.54532766342163, "per_seed": [ 53.77455520629883, 45.56505584716797, 42.8125, 40.686622619628906, 39.3796501159668, 41.54080581665039, 41.00275421142578, 51.600677490234375 ] }, { "cfg": { "lr": 0.003, "beta": 0.9, "tau": 0.1 }, "mean": 46.63834524154663, "per_seed": [ 54.938777923583984, 47.869895935058594, 44.608856201171875, 41.21122360229492, 42.689353942871094, 43.306400299072266, 44.728363037109375, 53.75389099121094 ] }, { "cfg": { "lr": 0.006, "beta": 0.9, "tau": 0.0 }, "mean": 21.585941553115845, "per_seed": [ 22.23833465576172, 17.9639835357666, 21.634672164916992, 20.441940307617188, 20.853069305419922, 23.09563446044922, 21.598304748535156, 24.86159324645996 ] }, { "cfg": { "lr": 0.006, "beta": 0.9, "tau": 0.05 }, "mean": 21.613163232803345, "per_seed": [ 22.23833465576172, 17.98293685913086, 21.707622528076172, 20.458215713500977, 20.899755477905273, 23.078847885131836, 21.629314422607422, 24.9102783203125 ] }, { "cfg": { "lr": 0.006, "beta": 0.9, "tau": 0.1 }, "mean": 22.09021544456482, "per_seed": [ 22.61715316772461, 18.391338348388672, 21.940710067749023, 20.666790008544922, 21.886329650878906, 23.460872650146484, 22.199935913085938, 25.55859375 ] } ], "custom_track": null }