Эксперимент: Model-Ensemble Space-Filling Explorer (#888)

{ "worked": false, "confidence": 9, "verdict": "Built a differentiable ensemble space-filling explorer with nonlinear candidate world models, log-determinant feature coverage, nominal and ensemble optimizers, finite-difference validation, ensemble-spread sweeps, and minibatch-gradient variance tests. The gradient check passed to 2.36e-9 maximum absolute error, and gradient MSE followed approximately 1/B with B*MSE near 4.8e-8. However, robust optimization did not improve worst-model log determinant: at spread 0.18 the change was -0.026%, far below the proposed 10% gain, so the central effect was not observed.", "metrics": { "baseline": "Nominal single-model design; worst-model Phi at spread 0.18 = 2.54541", "idea": "Ensemble design; worst-model Phi at spread 0.18 = 2.54475, gain = -0.026%; gradient MSE: 4.87e-8 (B=1), 7.48e-10 (B=64)" }, "how_to_run": "/home/maxwelhelp/main/bin/python3 ensemble_explorer.py", "files": [ "ensemble_explorer.py", "results.json", "REPORT.md", "article.md" ], "limitations": "Only a small open-loop scalar toy system and one fixed initialization were tested; no neural predictor training, learned policy, safety constraints, held-out prediction error, multiple random seeds, GPU execution, or larger model ensembles were evaluated." }