{ "bench_version": 1, "track": "tabular", "model": "mlp_tiny", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.006, "weight_decay": 0.0001 }, "sweep": [ { "cfg": { "lr": 0.001, "weight_decay": 0.0 }, "mean": 7.018276214599609 }, { "cfg": { "lr": 0.001, "weight_decay": 0.0001 }, "mean": 6.867959260940552 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 6.590747117996216 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0001 }, "mean": 6.44768226146698 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0 }, "mean": 5.89583158493042 }, { "cfg": { "lr": 0.006, "weight_decay": 0.0001 }, "mean": 5.276575446128845 } ], "full": { "mean": 5.42892998456955, "std": 0.41047416569067036, "per_seed": [ 5.578059196472168, 5.473832130432129, 5.200394153594971, 4.854016304016113, 6.361762046813965, 5.232422828674316, 5.255797863006592, 5.475155353546143 ], "n": 8 }, "grid_union": [ { "lr": 0.001, "weight_decay": 0.0 }, { "lr": 0.001, "weight_decay": 0.0001 }, { "lr": 0.003, "weight_decay": 0.0 }, { "lr": 0.003, "weight_decay": 0.0001 }, { "lr": 0.006, "weight_decay": 0.0 }, { "lr": 0.006, "weight_decay": 0.0001 } ] }, "idea": { "mean": 6.97602242231369, "std": 0.26847889510235706, "per_seed": [ 6.515726566314697, 6.685352802276611, 7.203780651092529, 7.1638712882995605, 6.83967924118042, 7.354576110839844, 6.9066338539123535, 7.138558864593506 ], "n": 8, "best_cfg": { "lr": 0.003, "delta": 0.5, "lambda": 0.03 } }, "comparison": { "delta_mean": 1.5470924377441406, "idea_wins": 0, "n_pairs": 8, "per_seed_diffs": [ 0.9376673698425293, 1.2115206718444824, 2.0033864974975586, 2.3098549842834473, 0.4779171943664551, 2.1221532821655273, 1.6508359909057617, 1.6634035110473633 ], "p_value": 0.0081, "mde": 0.5261925661705102, "mde_rel_pct": 9.692380776066154, "verdict": "idea worse (significant)", "system_worked": false }, "mechanism_signature": { "math_sanity": { "rows": [ { "delta": 2.0, "I_acq": 0.9397995625962782, "I_use": 0.05860868669192204 }, { "delta": 1.0, "I_acq": 1.5434491151928815, "I_use": 0.08289914666261401 }, { "delta": 0.5, "I_acq": 2.2116158746500787, "I_use": 0.09305665042940188 }, { "delta": 0.25, "I_acq": 2.899102061566953, "I_use": 0.09884338786391983 }, { "delta": 0.125, "I_acq": 3.589851585897336, "I_use": 0.10102097970960289 }, { "delta": 0.0625, "I_acq": 4.281741767709677, "I_use": 0.10177035755518295 }, { "delta": 0.03125, "I_acq": 4.97355053968472, "I_use": 0.10251417051178979 } ], "acq_log_slope": 0.9963986336397874, "acq_log_r2": 0.9999983290351192, "fine_acq_increment": 0.691808771975043, "fine_use_increment": 0.0007438129566068458 }, "idea_sweep": [ { "cfg": { "lr": 0.001, "delta": 0.5, "lambda": 0.03 }, "mean": 9.066823542118073, "per_seed": [ 11.198817253112793, 8.665545463562012, 9.605782508850098, 8.392980575561523, 7.9980645179748535, 9.001076698303223, 8.63749885559082, 9.034822463989258 ] }, { "cfg": { "lr": 0.001, "delta": 1.0, "lambda": 0.03 }, "mean": 56.89865565299988, "per_seed": [ 13.655369758605957, 14.693795204162598, 112.16630554199219, 9.177323341369629, 139.04576110839844, 13.599967002868652, 138.39633178710938, 14.454391479492188 ] }, { "cfg": { "lr": 0.001, "delta": 2.0, "lambda": 0.03 }, "mean": 27.349403142929077, "per_seed": [ 14.268783569335938, 11.69338607788086, 12.973438262939453, 11.941506385803223, 12.154528617858887, 11.955282211303711, 93.44461059570312, 50.36368942260742 ] }, { "cfg": { "lr": 0.003, "delta": 0.5, "lambda": 0.03 }, "mean": 6.97602242231369, "per_seed": [ 6.515726566314697, 6.685352802276611, 7.203780651092529, 7.1638712882995605, 6.83967924118042, 7.354576110839844, 6.9066338539123535, 7.138558864593506 ] }, { "cfg": { "lr": 0.006, "delta": 0.5, "lambda": 0.03 }, "mean": 7.816548943519592, "per_seed": [ 8.138494491577148, 8.348912239074707, 7.035199165344238, 6.89385461807251, 6.767551422119141, 7.8906731605529785, 7.620196342468262, 9.837510108947754 ] } ], "mechanism_signature": { "predicted": "finite-resolution channel suppresses noisy observations and rejects anti-aligned updates", "observed": { "accepted_fraction": 0.7734375, "mean_raw_grad_norm": 27.050026551771833, "mean_channel_update_norm": 27.05706130868445, "delta": 0.5 }, "confirmed": true } } }