{ "bench_version": 1, "track": "dynamics", "model": "rnn_small", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.006, "weight_decay": 0.0001 }, "sweep": [ { "cfg": { "lr": 0.0015, "weight_decay": 0.0 }, "mean": 0.005481100233737379 }, { "cfg": { "lr": 0.0015, "weight_decay": 0.0001 }, "mean": 0.005474042263813317 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 0.003975238680141047 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0001 }, "mean": 0.00394076865632087 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0 }, "mean": 0.002146739774616435 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0001 }, "mean": 0.002088654466206208 } ], "full": { "mean": 0.0022117759071988985, "std": 0.0004126881098400608, "per_seed": [ 0.002800536109134555, 0.001347046927548945, 0.001825663261115551, 0.0023813715670257807, 0.0023169058840721846, 0.0024658460170030594, 0.0022247303277254105, 0.002332107163965702 ], "n": 8 } }, "idea": { "mean": 0.0020463809778448194, "std": 0.0006329416993684799, "per_seed": [ 0.003160907654091716, 0.0019281062996014953, 0.0024306366685777903, 0.0027447089087218046, 0.0015351816546171904, 0.0011941296979784966, 0.0018756426870822906, 0.001501734252087772 ], "n": 8, "best_cfg": { "lr": 0.006, "weight_decay": 0.0001, "cert_weight": 0.01 }, "sweep": [ { "cfg": { "lr": 0.006, "weight_decay": 0.0001, "cert_weight": 0.01 }, "mean": 0.0025660898827482015 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0001, "cert_weight": 0.05 }, "mean": 0.0025660898827482015 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0001, "cert_weight": 0.2 }, "mean": 0.0025660898827482015 } ] }, "comparison": { "delta_mean": -0.00016539492935407907, "idea_wins": 4, "n_pairs": 8, "per_seed_diffs": [ 0.00036037154495716095, 0.0005810593720525503, 0.0006049734074622393, 0.00036333734169602394, -0.0007817242294549942, -0.0012717163190245628, -0.0003490876406431198, -0.0008303729118779302 ], "p_value": 0.54535, "mde": 0.0006149647377560715, "mde_rel_pct": 27.80411594838706, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "predicted_vs_observed": { "n_test": 800, "predicted_certificate_drop_mean": 0.3020721338689327, "observed_certificate_drop_mean": -0.014734884549397975, "predicted_vs_observed_drop_ratio": -0.048653354892595646, "contract_permitted_fraction": 0.8, "permitted_certificate_violations": 0, "confirmed": 1.0 }, "confirmed": 1.0 } }