Graph-Certified Switching SSM / report_bench_2026-08-31T113157.md
Mechanism confirmed, baseline not beaten
Стенд-проверка (stage-2) · промт оператора:
(универсальный)
Ответ агента:
{ "worked": false, "confidence": 9, "verdict": "Implemented Graph-Certified Switching SSM on the structurally matched dynamics/control track using the same four-mode recurrent architecture for baseline and idea, with the all-mode quadratic contraction loss as the only intervention. The trained-model signature confirmed the observed maximum zero-input certificate ratio was 0.574 versus target rho=0.81, but test MSE worsened from 0.06002 to 0.06697, with paired delta +0.00695 and permutation p=0.23195. Therefore the stability signal was present, but there was no task-metric win.", "metrics": { "baseline": "best lr=0.01; test MSE mean 0.0600216757, std 0.0151569244; per-seed [0.08692939, 0.06139367, 0.07089250, 0.05957052, 0.04860168, 0.06793582, 0.05194413, 0.03290570]", "idea": "best lambda_contract=0.2 at lr=0.01; test MSE mean 0.0669692494, std 0.0127851227; per-seed [0.07059102, 0.05721509, 0.08217970, 0.06861463, 0.06553272, 0.06471398, 0.08505756, 0.04184929]", "comparison": "delta_mean=+0.0069475737, idea_wins=3/8, permutation p_value=0.23195, verdict=no significant win", "mechanism_signature": "predicted contraction ratio <= 0.81; observed maximum zero-input ratio 0.5740312, observed mean ratio 0.1977006, confirmed=true across 8 trained models" }, "how_to_run": "/home/maxwelhelp/main/bin/python3 bench_experiment.py", "files": [ "bench_experiment.py", "bench_report.json" ], "bench_report": { "bench_version": 1, "track": "dynamics", "model": "switching_rnn_small", "metric_direction": "lower is better", "n_seeds": 8, "baseline": { "best_cfg": { "lr": 0.01, "weight_decay": 0.0 }, "sweep": [ { "cfg": { "lr": 0.001, "weight_decay": 0.0 }, "mean": 0.6787904053926468 }, { "cfg": { "lr": 0.003, "weight_decay": 0.0 }, "mean": 0.33453836664557457 }, { "cfg": { "lr": 0.01, "weight_decay": 0.0 }, "mean": 0.06969651952385902 } ], "full": { "mean": 0.060021675657480955, "std": 0.015156924357368727, "per_seed": [ 0.08692938834428787, 0.06139367073774338, 0.07089249789714813, 0.059570521116256714, 0.048601675778627396, 0.06793582439422607, 0.051944129168987274, 0.0329056978225708 ], "n": 8 } }, "idea": { "mean": 0.06696924939751625, "std": 0.012785122715736058, "per_seed": [ 0.0705910176038742, 0.057215094566345215, 0.08217969536781311, 0.06861463189125061, 0.06553272157907486, 0.0647139772772789, 0.0850575640797615, 0.0418492928147316 ], "n": 8 }, "comparison": { "delta_mean": 0.0069475737400352955, "idea_wins": 3, "n_pairs": 8, "per_seed_diffs": [ -0.016338370740413666, -0.004178576171398163, 0.011287197470664978, 0.009044110774993896, 0.016931045800447464, -0.003221847116947174, 0.03311343491077423, 0.008943594992160797 ], "p_value": 0.23195, "mde": 0.012566084314756818, "mde_rel_pct": 20.935910530832725, "verdict": "no significant win", "system_worked": false }, "mechanism_signature": { "signature": { "certificate": "V(z)=||z||^2; one-node path-complete graph; four all-mode self-edges", "prediction": "edge contraction ratio <= rho", "rho": 0.81, "observed_max_zero_input_ratio": 0.5740312337875366, "observed_mean_ratio": 0.19770055771505213, "confirmed": true, "n_trained_models": 8 }, "idea_cfg": { "lr": 0.01, "lambda_contract": 0.2 }, "training_budget": "400 train/400 test samples, 5 epochs, batch 128; equal on both systems" }, "idea_sweep": [ { "cfg": { "lr": 0.01, "lambda_contract": 0.01 }, "result": { "mean": 0.06696925032883883, "std": 0.012785122979561253, "per_seed": [ 0.0705910250544548, 0.057215094566345215, 0.08217969536781311, 0.06861463189125061, 0.06553272157907486, 0.0647139772772789, 0.0850575640797615, 0.0418492928147316 ], "n": 8 } }, { "cfg": { "lr": 0.01, "lambda_contract": 0.05 }, "result": { "mean": 0.06696925032883883, "std": 0.012785122979561253, "per_seed": [ 0.0705910250544548, 0.057215094566345215, 0.08217969536781311, 0.06861463189125061, 0.06553272157907486, 0.0647139772772789, 0.0850575640797615, 0.0418492928147316 ], "n": 8 } }, { "cfg": { "lr": 0.01, "lambda_contract": 0.2 }, "result": { "mean": 0.06696924939751625, "std": 0.012785122715736058, "per_seed": [ 0.0705910176038742, 0.057215094566345215, 0.08217969536781311, 0.06861463189125061, 0.06553272157907486, 0.0647139772772789, 0.0850575640797615, 0.0418492928147316 ], "n": 8 } } ], "custom_track": null, "device": "cuda" }, "limitations": "The final run used a reduced 400-sample train/test split and 5 epochs because the initial larger implementation exceeded the 40-minute limit. It tested a one-node quadratic certificate rather than the full multi-node reachability-envelope construction, and did not evaluate long-horizon adversarial switching rollouts or compare against the canonical bench rnn_small GRU architecture.", "system_verdict": "failed", "practical_verdict": "inconclusive", "mechanism_ok": 0, "system_judged": true }