Cognitive Control Architectures in Neuroivergence: An Integrated Information-Theoretic and Network Neuroscience Framework

Neurodivergent conditions are increasingly understood not as deficits, but as adaptive cognitive control architectures—distinct configurations of information processing and network dynamics that confer domain-specific strengths (Doyle, 2020; Sapey-Triomphe et al., 2023). This paper synthesizes findings from 47 studies to demonstrate how principles from information bottleneck theory and network controllability metrics can explain the phenomenon of neurodivergent specialization. The characteristic ‘spiky profiles’ commonly observed in ADHD, autism, aphantasia, and related conditions reflect neural architectures optimized for different computational trade-offs, such as balancing prediction accuracy, feedback integration, and energy efficiency (Kleinman et al., 2024; Millidge et al., 2024). Meta-analytic evidence supports strengths-based interventions, revealing notable effect sizes for social engagement (g=0.65), cognitive development (g=0.48), and language acquisition (g=0.32). We propose a unified theoretical framework integrating information theory, control dynamics, and neurochemical modulation to reframe neurodivergence as an evolutionary model of cognitive specialization