Sleep and quiet wakefulness signify an idling brain hub for creative insights

The review addresses how sleep and quiet wakefulness (idling states) contribute to memory processing and higher-order cognition, including creativity, insight, rule learning and problem-solving. Building on Hebbian plasticity and long-term potentiation, the authors frame memory as engrams—neuronal assemblies formed during experience, reactivated during post-learning sleep for consolidation, and later reinstated during recall. They highlight that idling states (sleep and quiet immobility) feature active neural processing comparable to online wakefulness in some regions (e.g., prefrontal cortex) and are critical in a two-stage memory model, where offline states follow exploratory behavior to stabilize and reorganize memories. The central aim is to synthesize recent evidence on offline reactivations and to explore links between sleep stages (NREM, REM), quiet wakefulness, and creative cognitive operations such as abstraction, inference, and insight.

The paper reviews foundational and contemporary literature on: (1) LTP and engrams as substrates for memory storage and reactivation; (2) the role of sleep in memory consolidation via hippocampal-cortical replays during NREM (sharp wave ripples) and REM; (3) systems-level findings that disrupting NREM ripples impairs spatial memory and that specific REM mechanisms (e.g., septal GABAergic modulation) support contextual engrams; (4) cortical top-down influences during NREM that consolidate perceptual memory; (5) ensemble-level calcium imaging showing orchestrated sub-ensembles within engrams, with about half of sub-ensembles reactivated during subsequent sleep; and (6) short-term memory existing as silent engrams that can be rendered naturally retrievable after facilitated consolidation dependent on NREM CA1 activity. Beyond consolidation, the review synthesizes human and animal evidence that sleep benefits creative cognition: REM sleep correlates with improved category learning and substantially enhances creative association performance (≈40% improvement on remote associates test; ≈32% better anagram solving after REM awakening compared to NREM). Nocturnal sleep fosters discovery of hidden rules (insight), and NREM slow waves may replay awake learning to build indirect associations. Awake hippocampal sharp-wave ripples can provide mnemonic shortcuts supporting inferential relationships. Collectively, these works support complementary NREM/REM functions in gist extraction and novel association formation, with quiet wakefulness also implicated in inference.

As a narrative review, the article synthesizes diverse methodologies from the literature and highlights specific experimental paradigms illustrating idling-state mechanisms: - Longitudinal calcium imaging (miniature endoscopes) to track engram and non-engram neuronal ensemble activity across learning, sleep, and retrieval, revealing coordinated sub-ensemble reactivations during sleep that support consolidation. - Electrophysiology of hippocampal sharp-wave ripples during NREM, including causal disruptions (e.g., ripple suppression) to assess effects on spatial memory and ensemble pattern consolidation. - Targeted manipulations of REM-related circuits (e.g., selective inhibition of medial septum GABA neurons projecting to dorsal hippocampus) to test stage-specific contributions to contextual engram stabilization. - Human behavioral paradigms: remote associates tests and anagram tasks administered across sleep/wake conditions, with REM versus NREM awakenings to evaluate creativity and problem-solving flexibility; category learning tasks with sleep quantification. - Mouse deductive reasoning-based transitive inference task using hierarchical contexts (A > B > C > D > E). Training includes only adjacent pairs; inference is assessed on a never-experienced pair (B vs D). Investigations include: sleep deprivation versus normal sleep; optogenetic inhibition of anterior cingulate cortex (ACC) during NREM or REM versus wake; pathway-specific necessity of medial entorhinal cortex (MEC) input to ACC during REM; and miniscope calcium imaging of ACC showing training-pair coactivations during NREM and emergence of inference-related patterns peaking during post-training REM. - Mouse memory assimilation paradigm leveraging geometric context similarity. After fear conditioning in one context, assimilation to a geometrically similar (but not dissimilar) context is tested across a sleep interval. Causal tests include silencing ACC (but not prelimbic cortex) during post-learning sleep and selective disruption of labeled engram populations during the assimilation sleep window. Calcium imaging quantifies co-reactivation of event representations during idling states, showing higher co-reactivation for assimilated event pairs. These complementary methods illuminate how specific sleep stages and quiet wakefulness support consolidation, inference, and assimilation through stage- and circuit-specific reactivations.

- Sleep and quiet wakefulness (idling states) actively process memory traces rather than being passive states, with reactivation (replay) of engram sub-ensembles during NREM and REM supporting consolidation and later retrieval. - Disruption of hippocampal sharp-wave ripples during NREM impairs spatial memory and stabilizing recent ensemble patterns; top-down cortical inputs during NREM consolidate perceptual memories. - Engram ensembles comprise coordinated sub-ensembles; approximately half of engram sub-ensembles reactivate during subsequent sleep to support consolidation. - Short-term memory can exist as a silent engram and become naturally retrievable after facilitated consolidation reliant on NREM CA1 activity. - REM sleep preferentially boosts creative cognition: about 40% improvement on a remote associates task after REM compared with NREM/quiet rest, and 32% better anagram solving after REM awakening compared with NREM awakening. - NREM slow-wave properties correlate with forming indirect associations in relational tasks; awake sharp-wave ripples can provide mnemonic shortcuts for inferential reasoning. - In mice, sleep is necessary to infer the correct choice in a transitive inference task (B vs D) after training only on adjacent pairs. Optogenetic inhibition of ACC during NREM or REM (but not wake) abolishes inference; MEC→ACC input during REM is required. Training-pair patterns coactivate in NREM, while inference-related patterns emerge and peak during post-training REM. - Memory assimilation across geometrically similar contexts requires sleep and ACC activity during the offline period; selective disruption of original engram reactivation during sleep impairs assimilation without erasing the original memory. Co-reactivation of representations for assimilated events is elevated during idling states. - Complementary roles: NREM supports reorganization, gist extraction, and schema integration; REM favors abstraction of indirect relationships and creative insight. Quiet wakefulness contributes to inference formation.

The reviewed evidence converges on the view that idling brain states are computationally active hubs that reorganize and integrate experiences to generate new knowledge. NREM-associated ripples and slow waves stabilize and restructure engram assemblies, extract gist, and integrate new learning into existing schemas. REM facilitates associative recombination and emergence of novel links, supporting creativity and insight. Quiet wakefulness offers additional windows for replay and inference building. Together, these processes explain how offline reactivation can transform discrete episodes into abstract rules, shortcuts, and creative solutions, addressing the central question of how sleep and idling support higher cognition. The rodent experiments demonstrate causal roles for specific prefrontal circuits (ACC) and inputs (MEC→ACC) in inference and assimilation across sleep stages, aligning with human findings that REM enhances creative performance and sleep promotes rule discovery. These insights highlight the functional specialization and complementarity of idling states in service of memory-based reasoning, problem-solving, and planning.

Sleep and quiet wakefulness comprise an idling brain mode that is indispensable for memory consolidation and for higher-order cognition. NREM sleep supports gist extraction, schema integration, and reorganization of learned content, while REM sleep promotes abstraction of indirect relationships and creative insights; quiet wakefulness aids inference. Emerging technological advances (miniscope imaging, precise optogenetic/chemogenetic manipulations, and high-resolution electrophysiology) have clarified stage- and circuit-specific mechanisms underpinning replay and reactivation. Future research should: (1) systematically dissect contributions of discrete sleep substages (e.g., N1 as a potential creative sweet spot) and quiet rest; (2) bridge species differences across rodents, non-human primates, and humans in sleep architecture and staging; (3) elucidate molecular, cellular, and ensemble-level mechanisms of idling-driven abstraction; and (4) extend investigations to complex cognitive domains including planning, decision-making, artistry, and problem-solving.

As a narrative review, the work synthesizes heterogeneous studies with varying methodologies, species, and sleep staging criteria, which may limit direct comparability and causal inference. Many human findings are correlational, and translating rodent circuit-level manipulations to humans is constrained by species differences in sleep architecture and prefrontal organization. The precise functional segregation among NREM substages and REM features remains incompletely characterized, and molecular/cellular mechanisms linking specific reactivation patterns to creative cognition are not fully understood.