Эксперимент: Covariance-Conditioned Neural Rollouts (#177)

{ "worked": true, "confidence": 8, "verdict": "Built a covariance-conditioned joint rollout MVP in mvp.py, including a Schur-complement PSD check and held-out AR(1) rollout comparison. The conditional head achieved nearly identical mean MSE to recursive prediction (2.72084 vs 2.72124), but substantially improved Gaussian NLL (14.23 vs 18.92) and 95% coverage (94.6% vs 86.9%) using the full correlated covariance. The effect is real for uncertainty quality, not a demonstrated mean-accuracy gain; redundant older prefix history did not materially help after observing the latest state.", "metrics": { "baseline": "Recursive AR rollout: MSE 2.72124, NLL 18.9187, 95% coverage 0.8687", "idea": "Joint covariance-conditioned rollout: MSE 2.72084, NLL 14.2336, 95% coverage 0.9458, conditional covariance trace 26.8955, minimum eigenvalue 0.2972; Schur complement minimum eigenvalue 0.2894" }, "how_to_run": "/home/maxwelhelp/main/bin/python3 mvp.py", "files": [ "mvp.py", "mvp_results.json", "experiment.py", "verify_mechanism.py", "mechanism_results.json", "results.json" ], "limitations": "The mini-experiment uses a stationary Gaussian AR(1) process rather than a trained GRU, transformer, Lorenz-63 system, or neural latent residual head. It evaluates a fixed horizon of 10 and does not measure FLOPs, runtime, low-rank covariance scaling, or performance under distribution shift." }