{ "bench_version": 1, "track": "dynamics", "model": "rnn_small", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.01 }, "sweep": [ { "cfg": { "lr": 0.001 }, "mean": 0.0010981351661030203 }, { "cfg": { "lr": 0.003 }, "mean": 0.0005773879893240519 }, { "cfg": { "lr": 0.01 }, "mean": 0.00017333678806608077 } ], "full": { "mean": 0.0001381416768708732, "std": 8.561204737527774e-05, "per_seed": [ 0.00027561953174881637, 0.00025423229089938104, 6.920921441633254e-05, 9.428611519979313e-05, 0.00011270481627434492, 4.6381595893763006e-05, 5.4802185331936926e-05, 0.00019789766520261765 ], "n": 8 } }, "idea": { "mean": 4.836640596295183e-05, "std": 1.6373566749923268e-05, "per_seed": [ 2.6542324121692218e-05, 3.756738078664057e-05, 8.299377805087715e-05, 3.5111366742057726e-05, 4.390031972434372e-05, 5.7259032473666593e-05, 4.6545013901777565e-05, 5.70120319025591e-05 ], "n": 8 }, "comparison": { "delta_mean": -8.977527090792137e-05, "idea_wins": 6, "n_pairs": 8, "per_seed_diffs": [ -0.00024907720762712415, -0.00021666491011274047, 1.3784563634544611e-05, -5.9174748457735404e-05, -6.88044965500012e-05, 1.0877436579903588e-05, -8.25717143015936e-06, -0.00014088563330005854 ], "p_value": 0.058, "mde": 8.536922866820681e-05, "mde_rel_pct": 61.79831503566081, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "prediction": "stable transverse perturbations should contract with slope near log(1+dt*alpha)/dt", "trained_model": "orbital phase RNN, seed 0, best benchmark lr", "predicted_transverse_alpha": -0.003641861502655863, "predicted_euler_slope": -0.003642856601714171, "observed_perturbation_slope": -0.004562072528716771, "max_edge_angle_rad": 0.2595794200897217, "confirmed": true }, "idea_sweep": [ { "cfg": { "lr": 0.001 }, "result": { "mean": 0.5228138975799084, "std": 0.05846373973675471, "per_seed": [ 0.46808767318725586, 0.4788855016231537, 0.5357304811477661, 0.6011213660240173, 0.5628133416175842, 0.42048752307891846, 0.5861251354217529, 0.5292601585388184 ], "n": 8 } }, { "cfg": { "lr": 0.003 }, "result": { "mean": 0.05459719488862902, "std": 0.024955650155415794, "per_seed": [ 0.03722333908081055, 0.034839458763599396, 0.0486103855073452, 0.06742417067289352, 0.08550677448511124, 0.015370250679552555, 0.09451965242624283, 0.05328352749347687 ], "n": 8 } }, { "cfg": { "lr": 0.01 }, "result": { "mean": 4.836640596295183e-05, "std": 1.6373566749923268e-05, "per_seed": [ 2.6542324121692218e-05, 3.756738078664057e-05, 8.299377805087715e-05, 3.5111366742057726e-05, 4.390031972434372e-05, 5.7259032473666593e-05, 4.6545013901777565e-05, 5.70120319025591e-05 ], "n": 8 } } ], "protocol": { "epochs": 12, "n_train": 1200, "n_test": 400, "paired_seeds": [ 0, 1, 2, 3, 4, 5, 6, 7 ], "lr_union": [ 0.001, 0.003, 0.01 ], "baseline_best_lr": 0.01, "idea_best_lr": 0.01 } }