Эксперимент: Adaptive reset neural ODE (#781)

{ "worked": true, "confidence": 8, "verdict": "Built a readable adaptive-reset neural ODE MVP with discrete 95th-percentile error-triggered boundaries, teacher resets during training, and predicted-state resets during deployment. The math mechanism manifested: tolerance boundaries matched analytic crossings within 0.00155 time units at dt=0.002, mismatch produced earlier boundaries monotonically, and tolerance produced longer windows monotonically. The toy adaptive rollout reduced RMSE substantially, but this is not a compute- or parameter-matched win because it used 166 local windows.", "metrics": { "baseline": "Shared scalar field: full-rollout RMSE 37.2274; final RMSE 85.4212.", "idea": "Adaptive local fields: full-rollout RMSE 2.4972; final RMSE 5.5191; 166 windows. Analytic-versus-grid boundary maximum absolute error 0.00155." }, "how_to_run": "/home/maxwelhelp/main/bin/python3 adaptive_reset_experiment.py", "files": [ "adaptive_reset_experiment.py", "results.json", "README.md" ], "limitations": "Only a scalar synthetic time-varying flow was tested; no neural network vector fields, chaotic oscillator, latent ODE, noisy observations, equal-parameter/FLOP budget, or GPU benchmark was evaluated. The adaptive result benefits from many local parameters and frequent reset opportunities." }