Robust Quasi-Conservation in Correction-Driven Information Networks: The Karishy Temporal–Drift Invariant (KTDI)

This work presents a computational proof-of-concept demonstrating the emergence of a robust quasi-conserved quantity in correction-driven abstract information networks. We introduce the Karishy Temporal–Drift Invariant (KTDI), defined as: KTDI (t) = ET (t) + b ES (t) + c EF (t) where ET denotes mean temporal-memory load, ES structural drift load, and EF correction residual (off-diagonal). Across systematic robustness sweeps over network size (N=30–80), drift coupling (γ=0. 10–0. 20), and temporal retention (ρ=0. 97–0. 99), the relative fluctuation of KTDI remained within 0. 50%–1. 57%, indicating strong quasi-conservation. The implementation provided represents a reduced demonstrative model designed to isolate the invariant-detection mechanism. Certain extended theoretical layers and architectural components are intentionally omitted.