Cross-Drive Toy Model Prediction in Selector-Time Theory: Persistence of Critical Memory Zones under Different Drives

This technical note presents the third stage of the computational sequence associated with Selector-Time Theory (STT), continuing previous work on a minimal irreversibility memory kernel, post-IRREV contextual typification, refined parameter sweeping, local robustness, and fine-grid stress testing. The article focuses on Listing 07, a cross-drive toy-model prediction designed to test whether the critical memory zone observed in Listings 03B–06 remains recognizable when the excitation form of the system is changed. Six synthetic drives are compared: triangular, sinusoidal, asymmetric ramp, random walk, piecewise shocks, and hybrid. The reference execution uses four controlled noise levels, fifty-one memory weights, and eight repetitions per drive-noise combination, totaling 9,792 kernel executions. The results indicate that the qualitative regime architecture is preserved across all tested drives: SUBMEMORIAL → CRITICAL_MEMORY → POST_CRITICAL_DECAY → LOCKED_MEMORY. The work does not claim direct physical validation of Selector-Time Theory. Instead, it presents a local, reproducible, and criticizable computational toy model showing that a hypothesis derived from STT can generate a simple prediction, expose systematic error, and remain transparent under controlled perturbations. Supplemental analyses include the observed critical-zone width by drive and the peak prediction error by noise level and drive. The publication is accompanied by the Python script, bilingual READMEs, CSV files, JSON summary, NPZ arrays, PNG figures, and terminal logs required for reproduction and technical auditing.