{ "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": 0.001222553226398304 }, { "cfg": { "lr": 0.003 }, "mean": 0.000683717749780044 }, { "cfg": { "lr": 0.01 }, "mean": 0.00021919916616752744 } ], "full": { "mean": 0.0002364131532885949, "std": 0.00010524220575938558, "per_seed": [ 0.00023485766723752022, 0.0002715944719966501, 0.00015523760521318763, 0.0002151069202227518, 0.00015697270282544196, 0.00027127040084451437, 0.0001109737713704817, 0.0004752916865982115 ], "n": 8 }, "idea_grid": [ { "cfg": { "lr": 0.001 }, "mean": 0.035339713853318244 }, { "cfg": { "lr": 0.003 }, "mean": 0.008510864296113141 }, { "cfg": { "lr": 0.01 }, "mean": 0.0002731473405219731 } ] }, "idea": { "mean": 0.0002731473405219731, "std": 0.00011425463363936107, "per_seed": [ 0.00023485766723752022, 0.0002715944719966501, 0.00015523760521318763, 0.00041881523793563247, 0.0002471378829795867, 0.00027127040084451437, 0.0001109737713704817, 0.0004752916865982115 ], "n": 8 }, "comparison": { "delta_mean": 3.673418723337818e-05, "idea_wins": 0, "n_pairs": 8, "per_seed_diffs": [ 0.0, 0.0, 0.0, 0.00020370831771288067, 9.016518015414476e-05, 0.0, 0.0, 0.0 ], "p_value": 0.5022, "mde": 6.227732731962294e-05, "mde_rel_pct": 26.342581389115686, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "mechanism_signature": { "prediction": "plateau-triggered delay amplifies parameter displacement", "predicted_delay_steps": 14.049755225604105, "observed_delay_steps": 12, "baseline_max_displacement": 6.033182621002197, "idea_max_displacement": 6.033182621002197, "displacement_ratio": 0.9999999999998342, "k_proxy": 0.1229826378985967, "confirmed": false } }, "custom_track": null }