A Network-Based Computational Model of Human Time Perception: Integrating Stress, Information Flow, and Sleep-Wake Differences

This work proposes a unified, network-based computational model of human time perception, integrating stress, information flow, and sleep–wake differences into a single mathematical framework. Rather than assuming a dedicated internal clock, the model formalizes subjective time as the integral of processing load across three major neural systems: prefrontal evaluative processing, hippocampal contextual updating, and amygdala-driven salience tagging. The paper synthesizes evidence from large-scale brain networks, predictive processing, time-cell physiology, and sleep neurobiology, and shows how these mechanisms jointly shape the expansion or compression of subjective time. The model explains why stress dilates time, why flow states compress it, and why long periods of sleep are experienced as instantaneous. A mathematically explicit formulation is provided, along with an implementation pathway for artificial cognitive architectures using internal processing load as a surrogate for subjective time. The work also outlines empirical predictions and validation strategies spanning behavioral experiments, neuroimaging, and artificial agents. Overall, the paper offers a unifying theoretical and computational account of time perception that bridges neuroscience and implementable AI systems.