Стенд-проверка (stage-2) · промт оператора:

(универсальный)

Ответ агента:

{ "worked": false, "confidence": 9, "verdict": "Implemented an end-to-end invariant-domain projection around the matched rnn_small dynamics model and ran the full 8-seed paired protocol with learning-rate sweep parity. The idea slightly improved test MSE, but the paired permutation test was not significant (delta_mean=-9.687e-06, p=0.0823), so there is no demonstrated win. The trained-model signature confirmed zero observed domain violations, but the limiter never activated because raw predictions already stayed inside the chosen angle domain.", "metrics": { "baseline": "Best lr=0.01; full 8-seed test MSE mean=0.0003015457559740753, std=0.00019741488644742065.", "idea": "Best lr=0.01; full 8-seed test MSE mean=0.0002918587179010501, std=0.00018751789044513985; 6/8 paired wins; delta_mean=-9.687038073025178e-06; permutation p=0.0823; limiter activation rate=0.0; raw and limited invalid-rate means=0.0." }, "how_to_run": "CUDA_VISIBLE_DEVICES='' /home/maxwelhelp/main/bin/python3 run_bench.py", "files": [ "run_bench.py", "bench_report.json", "bench_stdout.txt" ], "limitations": "The matched built-in dynamics track tests stability/control structure but is not a finite-volume 2-D Euler/PDE task, so density/pressure positivity was approximated by a scalar pendulum angle-domain projection. The limiter was not stress-tested on trained outputs because it never activated; no Euler solver, shock accuracy, geometry generalization, GPU run, or wall-clock/FLOP comparison was performed.", "bench_report": { "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": 0.0015202847716864198 }, { "cfg": { "lr": 0.003 }, "mean": 0.0009046953928191215 }, { "cfg": { "lr": 0.01 }, "mean": 0.00043617923438432626 } ], "full": { "mean": 0.0003015457559740753, "std": 0.00019741488644742065, "per_seed": [ 0.0006247494602575898, 0.0004462092474568635, 0.00017418812785763294, 0.0004995701019652188, 0.0003730663738679141, 0.00010612085316097364, 6.512444815598428e-05, 0.00012333743507042527 ], "n": 8 } }, "idea": { "best_cfg": { "lr": 0.01, "floor": 1e-05 }, "mean": 0.0002918587179010501, "std": 0.00018751789044513985, "per_seed": [ 0.0005810419097542763, 0.0004426969971973449, 0.00017546206072438508, 0.0005002013058401644, 0.00034365191822871566, 0.00010483426012797281, 6.512441905215383e-05, 0.00012185687228338793 ], "n": 8, "sweep": [ { "cfg": { "lr": 0.001, "floor": 1e-05 }, "mean": 0.0015202843060251325 }, { "cfg": { "lr": 0.003, "floor": 1e-05 }, "mean": 0.0009028831991599873 }, { "cfg": { "lr": 0.01, "floor": 1e-05 }, "mean": 0.00042485056837904267 } ] }, "comparison": { "delta_mean": -9.687038073025178e-06, "idea_wins": 6, "n_pairs": 8, "per_seed_diffs": [ -4.370755050331354e-05, -3.5122502595186234e-06, 1.2739328667521477e-06, 6.312038749456406e-07, -2.9414455639198422e-05, -1.2865930330008268e-06, -2.9103830456733704e-11, -1.4805627870373428e-06 ], "p_value": 0.0823, "mde": 1.4284899018237163e-05, "mde_rel_pct": 4.737224363212485, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "track_choice": "dynamics matches stability/control structure", "prediction": "hard convex interpolation eliminates out-of-domain reconstructed states", "raw_invalid_rate_mean": 0.0, "limited_invalid_rate_mean": 0.0, "activation_rate_mean": 0.0, "confirmed": true } }, "system_verdict": "partial", "practical_verdict": "inconclusive", "mechanism_ok": 1, "system_judged": true }