Quantum Logic Cellular Automata as Reservoirs: Unveiling the Information-Processing Role of Class IV

Abstract Reservoir computing (RC) achieves high computational performance near critical regimes, often termed the edge of chaos. However, inducing and maintaining such dynamics typically requires careful parameter tuning, limiting robustness and applicability. We propose a tuning-free critical reservoir architecture based on a Quantum-Logic Cellular Automaton (QLCA), whose transition relations induce a non-distributive orthomodular lattice via rough-set fixed points. Here, “quantum-logical” denotes a non-distributive information structure, not a quantum physical system.Using the five-bit memory task with distractors, QLCA reservoirs consistently achieve high success rates across diverse rules and outperform elementary cellular automata (ECA) reservoirs, particularly for intermediate and long distractors. Rules exhibiting Class-IV-like patterns maintain robust performance even at distractor lengths up to \(T_d=1000\), where conventional CA reservoirs rapidly degrade.By decomposing these dynamics into periodic backgrounds and propagating foreground structures, we show that long-horizon robustness depends on their interaction rather than foreground criticality alone. This interaction sustains information retention while mitigating distractor-induced degradation.These findings suggest that non-distributive quantum-logical structure promotes critical dynamics and enhances reservoir robustness without parameter tuning, offering a new design principle for RC with potential applications to physical and bio-inspired computational substrates.