Эксперимент: Encoder-reset recursive world-model training (#433)

{ "worked": true, "confidence": 9, "verdict": "Built an encoder-reset recursive world-model MVP and verified the core contraction mechanism with parameter sweeps. For every tested rho, observed log-decay slopes and 20-step perturbation ratios matched log(rho) and rho^20 to numerical precision, including the predicted rho=1 transition from decay to growth. The training comparison did not show a consistent performance win for reset versus carried state, so the stability mechanism is confirmed but the practical advantage remains unproven.", "metrics": { "baseline": "Carried-state model final-20-batch MSE: 0.003115 (N=16), 0.002179 (N=64).", "idea": "Encoder-reset model final-20-batch MSE: 0.005981 (N=16), 0.002257 (N=64). Exact toy sweeps matched predicted log(rho) slopes and rho^20 ratios for rho from 0.5 to 1.2." }, "how_to_run": "/home/maxwelhelp/main/bin/python3 experiment.py", "files": [ "experiment.py", "results.json", "REPORT.txt" ], "limitations": "The training comparison used one random seed, a small oscillator, short training, and a mean-pooled four-step context. It did not test GRU-TBPTT, N=256, L=16/64, changing dynamics, spectral normalization, Jacobian penalties, diminishing learning rates, FLOPs, or multiple seeds." }