REM Sleep as a Cognitive and Emotional Simulator for Decision-Making under Uncertainty: A Hypothesis/Theoretical Framework

In a dynamic and uncertain world, the human brain must efficiently process emotions and simulate potential scenarios to enhance decision-making under ambiguity. This theoretical framework proposes that rapid eye movement (REM) sleep serves as a cognitive simulator, decoding unconscious emotions and incoming information within a safe, virtual environment. During REM sleep, limbic system structures—such as the amygdala, hippocampus, and insula integrate recent sensory inputs with long-term memory schemas, transforming raw emotional signals based on potential threat and reward into actionable insights. Simultaneously, REM sleep facilitates affect-centric probabilistic simulations through hippocampal-prefrontal interactions, generating symbolic reconstructions of past experiences and hypothetical scenarios to foster cognitive flexibility and resilience. This model extends traditional perspectives on REM sleep, which emphasize memory consolidation and emotional regulation, by highlighting its critical role in future-oriented decision-making. Drawing on insights from neuroscience, psychology, and computational analogies with artificial intelligence, this framework underscores REM sleep’s dual function in emotional familiarization and adaptive strategy rehearsal. Recent evidence from studies on sleep deprivation and REM-specific manipulations further supports this, showing impairments in decision-making, insight generation, and risk assessment when REM is disrupted. Empirical gaps, including the need for direct neurobehavioral evidence linking REM to specific decision-making outcomes, and alternative hypotheses, such as the role of non-REM sleep or slow-wave activity, are discussed to contextualize the model’s contributions and guide future interdisciplinary research.