Category: Nonlinear Dyn.

Summary: Asking whether delayed adaptive release can reduce destructive avalanches in an OFC-style threshold system before control fatigue makes intervention counterproductive.

Threshold-driven systems such as earthquake-style slider models can alternate between long quiet periods and large cascade events. This experiment studies whether an adaptive release policy can improve resilience by bleeding off stress at the right moment, while also accounting for fatigue or diminishing effectiveness of repeated intervention.

The script relaxes avalanches, tests different release thresholds, and measures re-entrant gain, avalanche size, dissipation, and intervention frequency. The goal is not merely to lower activity, but to find whether there is a timing or threshold window where controlled release reduces large events without exhausting the control mechanism.

That window matters because real mitigation strategies are rarely free or infinitely repeatable. The experiment therefore probes whether adaptive intervention has a sweet spot between passivity and overuse.

Method: OFC-like avalanche simulations with adaptive release policies and threshold sweeps, aggregated across repeated trials under a fixed compute budget.

What is measured: Re-entrant gain, resilience, avalanche size, large-event fraction, intervention rate, dissipation, and best release threshold.