The function of dream sleep

Daydreaming Hopfield Networks and their surprising effectiveness on correlated data

To improve the storage capacity of the Hopfield model, we develop a version of the dreaming algorithm that perpetually reinforces the patterns to be stored (as in the Hebb rule), and erases the spurious memories (as in dreaming algorithms). For this reason, we called it Daydreaming. Daydreaming is not destructive and it converges asymptotically to stationary retrieval maps. When trained on random uncorrelated examples, the model shows optimal performance in terms of the size of the basins of attraction of stored examples and the quality of reconstruction. We also train the Daydreaming algorithm on correlated data obtained via the random-features model and argue that it spontaneously exploits the correlations thus increasing even further the storage capacity and the size of the basins of attraction. Moreover, the Daydreaming algorithm is also able to stabilize the features hidden in the data. Finally, we test Daydreaming on the MNIST dataset and show that it still works surprisingly well, producing attractors that are close to unseen examples and class prototypes.

Slumber under pressure: REM sleep and stress response

Sleep, a state of reduced responsiveness and distinct brain activity, is crucial across the animal kingdom. This review explores the potential adaptive functions of REM sleep in adapting to stress, emphasizing its role in memory consolidation, emotional regulation, and threat processing. We further explore the underlying neural mechanisms linking stress responses to REM sleep. By synthesizing current findings, we propose that REM sleep allows animals to "rehearse" or simulate responses to danger in a secure, offline state, while also maintaining emotional balance. Environmental factors, such as predation risk and social dynamics, further influence REM sleep. This modulation may enhance survival by optimizing stress responses while fulfilling physiological needs in animals. Insights into REM sleep's role in animals may shed light on human sleep in the context of modern stressors and sleep disruptions. This review also explores the complex interplay between stress, immunity, sleep disruptions-particularly involving REM sleep-and their evolutionary underpinnings.

Physical Considerations in Memory and Information Storage

Information is an important resource. Storing and retrieving information faithfully are huge challenges and many methods have been developed to understand the principles behind robust information processing. In this review, we focus on information storage and retrieval from the perspective of energetics, dynamics, and statistical mechanics. We first review the Hopfield model of associative memory, the classic energy-based model of memory. We then discuss generalizations and physical realizations of the Hopfield model. Finally, we highlight connections to energy-based neural networks used in deep learning. We hope this review inspires new directions along the lines of information storage and retrieval in physical systems.

Both slow wave and rapid eye movement sleep contribute to emotional memory consolidation

Sleep supports memory consolidation, but the specific roles of different sleep stages in this process remain unclear. While rapid eye movement sleep (REM) has traditionally been linked to the processing of emotionally charged material, recent evidence suggests that slow wave sleep (SWS) also plays a role in strengthening emotional memories. Here, we use targeted memory reactivation (TMR) during REM and SWS in a daytime nap to directly examine which sleep stage is primarily involved in consolidating emotional declarative memories. Contrary to our hypothesis, reactivating emotional stimuli during REM impairs memory. Meanwhile, TMR benefit in SWS is strongly correlated with the product of time spent in REM and SWS. The emotional valence of cued items modulates both delta/theta power and sleep spindles. Furthermore, emotional memories benefit more from TMR than neutral ones. Our findings suggest that SWS and REM have complementary roles in consolidating emotional memories, with REM potentially involved in forgetting them. These results also expand on recent evidence highlighting a connection between sleep spindles and emotional processing.

Learning in Associative Networks Through Pavlovian Dynamics

Hebbian learning theory is rooted in Pavlov's classical conditioning While mathematical models of the former have been proposed and studied in the past decades, especially in spin glass theory, only recently has it been numerically shown that it is possible to write neural and synaptic dynamics that mirror Pavlov conditioning mechanisms and also give rise to synaptic weights that correspond to the Hebbian learning rule. In this article we show that the same dynamics can be derived with equilibrium statistical mechanics tools and basic and motivated modeling assumptions. Then we show how to study the resulting system of coupled stochastic differential equations assuming the reasonable separation of neural and synaptic timescale. In particular, we analytically demonstrate that this synaptic evolution converges to the Hebbian learning rule in various settings and compute the variance of the stochastic process. Finally, drawing from evidence on pure memory reinforcement during sleep stages, we show how the proposed model can simulate neural networks that undergo sleep-associated memory consolidation processes, thereby proving the compatibility of Pavlovian learning with dreaming mechanisms.

Memory engram stability and flexibility

Many studies have shown that memories are encoded in sparse neural ensembles distributed across the brain. During the post-encoding period, often during sleep, many of the cells that were active during encoding are reactivated, supporting consolidation of this memory. During memory recall, many of the same cells that were active during encoding and reactivated during consolidation are reactivated during recall. These ensembles of cells have been referred to as the memory engram cells, stably representing a specific memory. However, recent studies question the rigidity of the "stable memory engram." Here we review the past literature of how episodic-like memories are encoded, consolidated, and recalled. We also highlight more recent studies (as well as some older literature) that suggest that these stable memories and their representations are much more dynamic and flexible than previously thought. We highlight some of these processes, including memory updating, reconsolidation, forgetting, schema learning, memory-linking, and representational drift.

Regularization, early-stopping and dreaming: A Hopfield-like setup to address generalization and overfitting

In this work we approach attractor neural networks from a machine learning perspective: we look for optimal network parameters by applying a gradient descent over a regularized loss function. Within this framework, the optimal neuron-interaction matrices turn out to be a class of matrices which correspond to Hebbian kernels revised by a reiterated unlearning protocol. Remarkably, the extent of such unlearning is proved to be related to the regularization hyperparameter of the loss function and to the training time. Thus, we can design strategies to avoid overfitting that are formulated in terms of regularization and early-stopping tuning. The generalization capabilities of these attractor networks are also investigated: analytical results are obtained for random synthetic datasets, next, the emerging picture is corroborated by numerical experiments that highlight the existence of several regimes (i.e., overfitting, failure and success) as the dataset parameters are varied.

Enhancing and advancing the understanding and study of dreaming and memory consolidation: Reflections, challenges, theoretical clarity, and methodological considerations

Empirical investigations that search for a link between dreaming and sleep-dependent memory consolidation have focused on testing for an association between dreaming of what was learned, and improved memory performance for learned material. Empirical support for this is mixed, perhaps owing to the inherent challenges presented by the nature of dreams, and methodological inconsistencies. The purpose of this paper is to address critically prevalent assumptions and practices, with the aim of clarifying and enhancing research on this topic, chiefly by providing a theoretical synthesis of existing models and evidence. Also, it recommends the method of Targeted Memory Reactivation (TMR) as a means for investigating if dream content can be linked to specific cued activations. Other recommendations to enhance research practice and enquiry on this subject are also provided, focusing on the HOW and WHY we search for memory sources in dreams, and what purpose (if any) they might serve.

How-tests for consciousness and direct neurophenomenal structuralism

Despite recent criticism, the search for neural correlates of consciousness (NCCs) is still at the core of a contemporary neuroscience of consciousness. One common aim is to distinguish merely statistical correlates from "NCCs proper", i.e., NCCs that are uniquely associated with a conscious experience and lend themselves to a metaphysical interpretation. We should then distinguish between NCCs as data and NCCs as hypotheses, where the first is just recorded data while the second goes beyond any set of recorded data. Still, such NCC-hypotheses ought to be testable. Here, I present a framework for so-called "sufficiency tests." We can distinguish four different classes of such tests, depending on whether they predict creature consciousness (which systems are conscious), state consciousness (when a system is conscious), phenomenal content (what a system is conscious of), or phenomenal character (how a system experiences). For each kind of test, I provide examples from the empirical literature. I also argue that tests for phenomenal character (How-Tests) are preferable because they bracket problematic aspects of the other kinds of tests. However, How-Tests imply a metaphysical tie between the neural and phenomenal domain that is stronger than supervenience, delivers explanations but does not close the explanatory gap, uses first-person methods to test hypotheses, and thereby relies on a form of direct neurophenomenal structuralism.

Hebbian dreaming for small datasets

The dreaming Hopfield model constitutes a generalization of the Hebbian paradigm for neural networks, that is able to perform on-line learning when "awake" and also to account for off-line "sleeping" mechanisms. The latter have been shown to enhance storing in such a way that, in the long sleep-time limit, this model can reach the maximal storage capacity achievable by networks equipped with symmetric pairwise interactions. In this paper, we inspect the minimal amount of information that must be supplied to such a network to guarantee a successful generalization, and we test it both on random synthetic and on standard structured datasets (i.e., MNIST, Fashion-MNIST and Olivetti). By comparing these minimal thresholds of information with those required by the standard (i.e., always "awake") Hopfield model, we prove that the present network can save up to ∼90% of the dataset size, yet preserving the same performance of the standard counterpart. This suggests that sleep may play a pivotal role in explaining the gap between the large volumes of data required to train artificial neural networks and the relatively small volumes needed by their biological counterparts. Further, we prove that the model Cost function (typically used in statistical mechanics) admits a representation in terms of a standard Loss function (typically used in machine learning) and this allows us to analyze its emergent computational skills both theoretically and computationally: a quantitative picture of its capabilities as a function of its control parameters is achieved and consistency between the two approaches is highlighted. The resulting network is an associative memory for pattern recognition tasks that learns from examples on-line, generalizes correctly (in suitable regions of its control parameters) and optimizes its storage capacity by off-line sleeping: such a reduction of the training cost can be inspiring toward sustainable AI and in situations where data are relatively sparse.

Sleep stage prediction using multimodal body network and circadian rhythm

Quality sleep plays a vital role in living beings as it contributes extensively to the healing process and the removal of waste products from the body. Poor sleep may lead to depression, memory deficits, heart, and metabolic problems, etc. Sleep usually works in cycles and repeats itself by transitioning into different stages of sleep. This study is unique in that it uses wearable devices to collect multiple parameters from subjects and uses this information to predict sleep stages and sleep patterns. For the multivariate multiclass sleep stage prediction problem, we have experimented with both memoryless (ML) and memory-based models on seven database instances, that is, five from the collected dataset and two from the existing datasets. The Random Forest classifier outclassed the ML models that are LR, MLP, kNN, and SVM with accuracy (ACC) of 0.96 and Cohen Kappa 0.96, and the memory-based model long short-term memory (LSTM) performed well on all the datasets with the maximum attained accuracy of 0.88 and Kappa 0.82. The proposed methodology was also validated on a longitudinal dataset, the Multiethnic Study of Atherosclerosis (MESA), with ACC and Kappa of 0.75 and 0.64 for ML models and 0.86 and 0.78 for memory-based models, respectively, and from another benchmarked Apple Watch dataset available on Physio-Net with ACC and Kappa of 0.93 and 0.93 for ML and 0.92 and 0.87 for memory-based models, respectively. The given methodology showed better results than the original work and indicates that the memory-based method works better to capture the sleep pattern.

Sleep loss disrupts decision-making ability and neuronal cytomorphology in zebrafish and the effects are mediated by noradrenaline acting on α1-adrenoceptor

Sleep is an instinct behavior, and its significance and functions are still an enigma. It is expressed throughout one's life and its loss affects psycho-somatic and physiological processes. We had proposed that it might maintain a fundamental property of the neurons and the brain. In that context, it was shown that sleep, rapid eye movement sleep (REMS) in particular, by regulating noradrenaline (NA), maintains the brain excitability. It was also reported that sleep-loss affected memory, reaction time and decision-making ability among others. However, as there was lack of clarity on the cause-and-effect relationship as to how the sleep-loss could affect these basic behaviors, their association was questioned and it was difficult to propose a cure or at least ways and means to ameliorate the symptoms. Also, we wanted to conduct the studies in a simpler model system so that conducting future molecular studies might be easier. Hence, using zebrafish as a model we evaluated if sleep-loss affected the basic decision-making ability, a cognitive process and if the effect was induced by NA. Indeed, our findings confirmed that upon sleep-deprivation, the cognitive decision-making ability of the prey zebrafish was compromised to protect itself by running away from the reach of the exposed predator Tiger Oscar (TO) fish. Also, we observed that upon sleep-loss the axonal arborization of the prey zebrafish brain was reduced. Interestingly, the effects were prevented by prazosin (PRZ), an α1-adrenoceptor (AR) antagonist and when the zebrafish recovered from the lost sleep.

Reduced left hippocampal perfusion is associated with insomnia in patients with cerebral small vessel disease

OBJECTIVES

Insomnia was associated with cerebral structural changes and Alzheimer's disease. However, associations among cerebral perfusion, insomnia with cerebral small vessel disease (CSVD), and cognitive performance were little investigated.

METHODS

This cross-sectional study included 89 patients with CSVDs and white matter hyperintensities (WMHs). They were dichotomized into the normal sleep and poor sleep group, according to Pittsburgh sleep quality index (PSQI). Baseline characteristics, cognitive performance, and cerebral blood flow (CBF) were measured and compared between the two groups. The association or correlation between cerebral perfusion, cognition, and insomnia was analyzed using binary logistic regression.

RESULTS

Our study found that declined MoCA score (P = .0317) was more prevalent in those with poor sleep. There was a statistical difference in the recall (P = .0342) of MMSE, the delayed recall (P = .0289) of MoCA between the two groups. Logistic regression analysis showed educational background (P < .001) and insomnia severity index (ISI) score (P = .039) were independently correlated with MoCA scores. Arterial spin labeling demonstrated that left hippocampal gray matter perfusion was significantly reduced (P = .0384) in the group with poor sleep. And, negative correlation was found between left hippocampal perfusion and PSQI scores.

CONCLUSIONS

In the patients with CSVDs, insomnia severity was associated with cognitive decline. Left hippocampal gray matter perfusion was correlated with PSQI scores in CSVDs.

Do sleep disruptions promote social fragmentation?

Sleep changes predate shifts in mood/affect, thought processing, mental and physical health, civic engagement, and contextual circumstances, among other things. Theory predicts that these changes may lead to shifts in political and social beliefs. Do sleep disruptions shape how individuals see the world, the people around them, and themselves in relation to others? In this article, we use daily survey data from the 77 waves (N ≈ 460,000) of the University of California, Los Angeles's 2019-2021 Nationscape Survey-a nationally representative political survey-to examine the effect of an exogenous short-term sleep disruption on measures of political views, polarization, and discriminatory beliefs. Using this data set, we leverage the modest sleep disruption that occurs at the start (and end) of Daylight Saving Time (DST) and employ a regression discontinuity in time design around the precise DST cutoff (which we supplement with event study models). Despite strong theoretical expectations and correlational connection between measures of sleep and many outcomes related to social fragmentation, we find that the DST change has little to no causal effect on citizens' levels of polarization or their discriminatory attitudes. These effects are precise enough to rule out small effects, robust to a host of specification checks, and consistent across potential subgroups of interest. Our work adds to a small but growing body of research on the social and political effects of sleep disruptions.

Musical hallucinations, secondary delusions, and lack of insight: results from a cohort study

Introduction

Although musical hallucinations do not tend to be accompanied by delusions, occasionally patients persistently accuse others of being responsible for causing the music they perceive, sometimes with severe social consequences such as frequently calling the police or moving house. In this study we seek to broaden our understanding of this rare type of musical hallucination that comes with secondary delusions and lack of insight, and to explore associations, underlying mechanisms, and treatment possibilities.

Methods

The present study is part of a cohort study on musical hallucinations carried out in the Netherlands from 2010 through 2023. Participants underwent testing with the aid of the MuHa Questionnaire, Launay-Slade Hallucinations Scale (LSHS), Schizotypal Personality Questionnaire (SPQ), Hamilton Depression Rating Scale (HDRS), and Mini Mental State Examination (MMSE). Additionally, they underwent a brain MRI, electroencephalogram, and audiological testing.

Results

Five patients out of a group of N = 81 (6%) lacked insight and presented with secondary delusions regarding the perceived music. They were all female, of advanced age, and hearing-impaired, and were diagnosed with cognitive impairment. In three patients (60%), risperidone was started. This had a positive effect on the hallucinations and secondary delusions.

Conclusion

The pathophysiological process underlying musical hallucinations is multifactorial in nature. We consider cognitive impairment the most likely contributing factor of the secondary delusions and lack of insight encountered in our patients, and antipsychotics the most beneficial treatment. On the basis of these small numbers, no definite conclusions can be drawn, so further research is needed to elucidate the underlying mechanisms and to develop evidence-based treatment methods for people experiencing this rare and debilitating combination of symptoms. Since the black box warning of risperidone cautions against the use of this drug in elderly persons with dementia, a proper comparison with the efficacy and safety of other antipsychotics for this group is paramount.

The Ponto-Geniculo-Occipital (PGO) Waves in Dreaming: An Overview

Rapid eye movement (REM) sleep is the main sleep correlate of dreaming. Ponto-geniculo-occipital (PGO) waves are a signature of REM sleep. They represent the physiological mechanism of REM sleep that specifically limits the processing of external information. PGO waves look just like a message sent from the pons to the lateral geniculate nucleus of the visual thalamus, the occipital cortex, and other areas of the brain. The dedicated visual pathway of PGO waves can be interpreted by the brain as visual information, leading to the visual hallucinosis of dreams. PGO waves are considered to be both a reflection of REM sleep brain activity and causal to dreams due to their stimulation of the cortex. In this review, we summarize the role of PGO waves in potential neural circuits of two major theories, i.e., (1) dreams are generated by the activation of neural activity in the brainstem; (2) PGO waves signaling to the cortex. In addition, the potential physiological functions during REM sleep dreams, such as memory consolidation, unlearning, and brain development and plasticity and mood regulation, are discussed. It is hoped that our review will support and encourage research into the phenomenon of human PGO waves and their possible functions in dreaming.

Targeted memory reactivation during slow-wave sleep vs. sleep stage N2: no significant differences in a vocabulary task

Sleep supports memory consolidation, and slow-wave sleep (SWS) in particular is assumed to benefit the consolidation of verbal learning material. Re-exposure to previously learned words during SWS with a technique known as targeted memory reactivation (TMR) consistently benefits memory. However, TMR has also been successfully applied during sleep stage N2, though a direct comparison between words selectively reactivated during SWS versus N2 is still missing. Here, we directly compared the effects of N2 TMR and SWS TMR on memory performance in a vocabulary learning task in a within-subject design. Thirty-four healthy young participants (21 in the main sample and 13 in an additional sample) learned 120 Dutch-German word pairs before sleep. Participants in the main sample slept for ∼8 h during the night, while participants in the additional sample slept ∼3 h. We reactivated the Dutch words selectively during N2 and SWS in one single night. Forty words were not cued. Participants in the main sample recalled the German translations of the Dutch words after sleep in the morning, while those in the additional sample did so at 2:00 a.m. As expected, we observed no differences in recall performance between words reactivated during N2 and SWS. However, we failed to find an overall memory benefit of reactivated over nonreactivated words. Detailed time-frequency analyses showed that words played during N2 elicited stronger characteristic oscillatory responses in several frequency bands, including spindle and theta frequencies, compared with SWS. These oscillatory responses did not vary with the memory strengths of individual words. Our results question the robustness and replicability of the TMR benefit on memory using our Dutch vocabulary learning task. We discuss potential boundary conditions for vocabulary reactivation paradigms and, most importantly, see the need for further replication studies, ideally including multiple laboratories and larger sample sizes.

Gallic Acid Ameliorates Cognitive Impairment Caused by Sleep Deprivation through Antioxidant Effect

Sleep deprivation (SD) has a profound impact on the central nervous system, resulting in an array of mood disorders, including depression and anxiety. Despite this, the dynamic alterations in neuronal activity during sleep deprivation have not been extensively investigated. While some researchers propose that sleep deprivation diminishes neuronal activity, thereby leading to depression. Others argue that short-term sleep deprivation enhances neuronal activity and dendritic spine density, potentially yielding antidepressant effects. In this study, a two-photon microscope was utilized to examine the calcium transients of anterior cingulate cortex (ACC) neurons in awake SD mice in vivo at 24-hour intervals. It was observed that SD reduced the frequency and amplitude of Ca2+ transients while increasing the proportions of inactive neurons. Following the cessation of sleep deprivation, neuronal calcium transients demonstrated a gradual recovery. Moreover, whole-cell patch-clamp recordings revealed a significant decrease in the frequency of spontaneous excitatory post-synaptic current (sEPSC) after SD. The investigation also assessed several oxidative stress parameters, finding that sleep deprivation substantially elevated the level of malondialdehyde (MDA), while simultaneously decreasing the expression of Nuclear Factor erythroid 2-Related Factor 2 (Nrf2) and activities of Superoxide dismutase (SOD) in the ACC. Importantly, the administration of gallic acid (GA) notably mitigated the decline of calcium transients in ACC neurons. GA was also shown to alleviate oxidative stress in the brain and improve cognitive impairment caused by sleep deprivation. These findings indicate that the calcium transients of ACC neurons experience a continuous decline during sleep deprivation, a process that is reversible. GA may serve as a potential candidate agent for the prevention and treatment of cognitive impairment induced by sleep deprivation.

Emotional memories are enhanced when reactivated in slow wave sleep, but impaired when reactivated in REM

Sleep supports memory consolidation. However, it is not completely clear how different sleep stages contribute to this process. While rapid eye movement sleep (REM) has traditionally been implicated in the processing of emotionally charged material, recent studies indicate a role for slow wave sleep (SWS) in strengthening emotional memories. Here, to directly examine which sleep stage is primarily involved in emotional memory consolidation, we used targeted memory reactivation (TMR) in REM and SWS during a daytime nap. Contrary to our hypothesis, reactivation of emotional stimuli during REM led to impaired memory. Consistent with this, REM% was correlated with worse recall in the group that took a nap without TMR. Meanwhile, cueing benefit in SWS was strongly correlated with the product of times spent in REM and SWS (SWS-REM product), and reactivation significantly enhanced memory in those with high SWS-REM product. Surprisingly, SWS-REM product was associated with better memory for reactivated items and poorer memory for non-reactivated items, suggesting that sleep both preserved and eliminated emotional memories, depending on whether they were reactivated. Notably, the emotional valence of cued items modulated both sleep spindles and delta/theta power. Finally, we found that emotional memories benefited from TMR more than did neutral ones. Our results suggest that emotional memories decay during REM, unless they are reactivated during prior SWS. Furthermore, we show that active forgetting complements memory consolidation, and both take place across SWS and REM. In addition, our findings expand upon recent evidence indicating a link between sleep spindles and emotional processing.

The stress of losing sleep: Sex-specific neurobiological outcomes

Sleep is a vital and evolutionarily conserved process, critical to daily functioning and homeostatic balance. Losing sleep is inherently stressful and leads to numerous detrimental physiological outcomes. Despite sleep disturbances affecting everyone, women and female rodents are often excluded or underrepresented in clinical and pre-clinical studies. Advancing our understanding of the role of biological sex in the responses to sleep loss stands to greatly improve our ability to understand and treat health consequences of insufficient sleep. As such, this review discusses sex differences in response to sleep deprivation, with a focus on the sympathetic nervous system stress response and activation of the hypothalamic-pituitary-adrenal (HPA) axis. We review sex differences in several stress-related consequences of sleep loss, including inflammation, learning and memory deficits, and mood related changes. Focusing on women's health, we discuss the effects of sleep deprivation during the peripartum period. In closing, we present neurobiological mechanisms, including the contribution of sex hormones, orexins, circadian timing systems, and astrocytic neuromodulation, that may underlie potential sex differences in sleep deprivation responses.

Dream interpretation and empirical dream research - an overview of research findings and their connections with psychoanalytic dream theories

The paper confronts psychoanalytic dream theories with the findings of empirical dream research. It summarizes the discussion in psychoanalysis around the function of dreams (e.g. as the guardian of sleep), wish-fullfilment or compensation, whether there is a difference between latent and manifest content, etc. In empirical dream research some of these questions have been investigated and the results can provide clarifications for psychoanalytic theorizing. The paper provides an overview of empirical dream research and its findings, as well as of clinical dream research in psychoanalysis, which was mainly conducted in German-speaking countries. The results are used to discuss the major questions in psychoanalytic dream theories and points out some developments in contemporary approaches which have been influenced by these insights. As a conclusion the paper attempts to formulate a revised theory of dreaming and its functions, which combines psychoanalytic thinking with research.

Interhemispheric competition during sleep

Our understanding of the functions and mechanisms of sleep remains incomplete, reflecting their increasingly evident complexity1-3. Likewise, studies of interhemispheric coordination during sleep4-6 are often hard to connect precisely to known sleep circuits and mechanisms. Here, by recording from the claustra of sleeping bearded dragons (Pogona vitticeps), we show that, although the onsets and offsets of Pogona rapid-eye-movement (REMP) and slow-wave sleep are coordinated bilaterally, these two sleep states differ markedly in their inter-claustral coordination. During slow-wave sleep, the claustra produce sharp-wave ripples independently of one another, showing no coordination. By contrast, during REMP sleep, the potentials produced by the two claustra are precisely coordinated in amplitude and time. These signals, however, are not synchronous: one side leads the other by about 20 ms, with the leading side switching typically once per REMP episode or in between successive episodes. The leading claustrum expresses the stronger activity, suggesting bilateral competition. This competition does not occur directly between the two claustra or telencephalic hemispheres. Rather, it occurs in the midbrain and depends on the integrity of a GABAergic (γ-aminobutyric-acid-producing) nucleus of the isthmic complex, which exists in all vertebrates and is known in birds to underlie bottom-up attention and gaze control. These results reveal that a winner-take-all-type competition exists between the two sides of the brain of Pogona, which originates in the midbrain and has precise consequences for claustrum activity and coordination during REMP sleep.

Sleep and dreaming in the light of reactive and predictive homeostasis

Dreams are often viewed as fascinating but irrelevant mental epihenomena of the sleeping mind with questionable functional relevance. Despite long hours of oneiric activity, and high individual differences in dream recall, dreams are lost into oblivion. Here, we conceptualize dreaming and dream amnesia as inherent aspects of the reactive and predictive homeostatic functions of sleep. Mental activity during sleep conforms to the interplay of restorative processes and future anticipation, and particularly during the second half of the night, it unfolds as a special form of non-constrained, self-referent, and future-oriented cognitive process. Awakening facilitates constrained, goal-directed prospection that competes for shared neural resources with dream production and dream recall, and contributes to dream amnesia. We present the neurophysiological aspects of reactive and predictive homeostasis during sleep, highlighting the putative role of cortisol in predictive homeostasis and forgetting dreams. The theoretical and methodological aspects of our proposal are discussed in relation to the study of dreaming, dream recall, and sleep-related cognitive processes.

Mammalian NREM and REM sleep: Why, when and how

This report proposes that fish use the spinal-rhombencephalic regions of their brain to support their activities while awake. Instead, the brainstem-diencephalic regions support the wakefulness in amphibians and reptiles. Lastly, mammals developed the telencephalic cortex to attain the highest degree of wakefulness, the cortical wakefulness. However, a paralyzed form of spinal-rhombencephalic wakefulness remains in mammals in the form of REMS, whose phasic signs are highly efficient in promoting maternal care to mammalian litter. Therefore, the phasic REMS is highly adaptive. However, their importance is low for singletons, in which it is a neutral trait, devoid of adaptive value for adults, and is mal-adaptive for marine mammals. Therefore, they lost it. The spinal-rhombencephalic and cortical wakeful states disregard the homeostasis: animals only attend their most immediate needs: foraging defense and reproduction. However, these activities generate allostatic loads that must be recovered during NREMS, that is a paralyzed form of the amphibian-reptilian subcortical wakefulness. Regarding the regulation of tonic REMS, it depends on a hypothalamic switch. Instead, the phasic REMS depends on an independent proportional pontine control.

Modeling integrated stress, sleep, fear and neuroimmune responses: Relevance for understanding trauma and stress-related disorders

Sleep and stress have complex interactions that are implicated in both physical diseases and psychiatric disorders. These interactions can be modulated by learning and memory, and involve additional interactions with the neuroimmune system. In this paper, we propose that stressful challenges induce integrated responses across multiple systems that can vary depending on situational variables in which the initial stress was experienced, and with the ability of the individual to cope with stress- and fear-inducing challenges. Differences in coping may involve differences in resilience and vulnerability and/or whether the stressful context allows adaptive learning and responses. We provide data demonstrating both common (corticosterone, SIH and fear behaviors) and distinguishing (sleep and neuroimmune) responses that are associated with an individual's ability to respond and relative resilience and vulnerability. We discuss neurocircuitry regulating integrated stress, sleep, neuroimmune and fear responses, and show that responses can be modulated at the neural level. Finally, we discuss factors that need to be considered in models of integrated stress responses and their relevance for understanding stress-related disorders in humans.

How to Represent Part-Whole Hierarchies in a Neural Network

This article does not describe a working system. Instead, it presents a single idea about representation that allows advances made by several different groups to be combined into an imaginary system called GLOM.1 The advances include transformers, neural fields, contrastive representation learning, distillation, and capsules. GLOM answers the question: How can a neural network with a fixed architecture parse an image into a part-whole hierarchy that has a different structure for each image? The idea is simply to use islands of identical vectors to represent the nodes in the parse tree. If GLOM can be made to work, it should significantly improve the interpretability of the representations produced by transformer-like systems when applied to vision or language.

Locus coeruleus and the defensive activation theory of rapid eye movement sleep: A mechanistic perspective

The defensive activation theory (DAT) was recently proposed to explain the biological function of dreaming. Briefly, DAT states that dreams are primarily visual to prevent plastic take-over of an otherwise inactive visual cortex during sleep. Evidence to support the DAT revolve around the interplay between dream activity (REM%) and cortical plasticity found in evolutionary history, primate studies, and coinciding decline in human cortical plasticity and REM% with age. As the DAT may prove difficult to test experimentally, we investigate whether further support for the DAT can be found in the literature. Plasticity and REM sleep are closely linked to functions of the Locus Coeruleus (LC). We therefore review existing knowledge about the LC covering LC stability with age, and the role of the LC in the plasticity of the visual cortex. Recent studies show the LC to be more stable than previously believed and therefore, the LC likely supports the REM% and plasticity in the same manner throughout life. Based on this finding, we review the effect of aging on REM% and visual cortex plasticity. Here, we find that recent, weighty studies are not in complete agreement with the data originally provided as support for DAT. Results from these studies, however, are not in themselves irreconcilable with the DAT. Our findings therefore do not disprove the DAT. Importantly, we show that the LC is involved in all mechanisms central to the DAT. The LC may therefore provide an experimental window to further explore and test the DAT.

Converging theories on dreaming: Between Freud, predictive processing, and psychedelic research

Dreams are still an enigma of human cognition, studied extensively in psychoanalysis and neuroscience. According to the Freudian dream theory and Solms' modifications of the unconscious derived from it, the fundamental task of meeting our emotional needs is guided by the principle of homeostasis. Our innate value system generates conscious feelings of pleasure and unpleasure, resulting in the behavior of approaching or withdrawing from the world of objects. Based on these experiences, a hierarchical generative model of predictions (priors) about the world is constantly created and modified, with the aim to optimize the meeting of our needs by reducing prediction error, as described in the predictive processing model of cognition. Growing evidence from neuroimaging supports this theory. The same hierarchical functioning of the brain is in place during sleep and dreaming, with some important modifications like a lack of sensual and motor perception and action. Another characteristic of dreaming is the predominance of primary process thinking, an associative, non-rational cognitive style, which can be found in similar altered states of consciousness like the effect of psychedelics. Mental events that do not successfully fulfill an emotional need will cause a prediction error, leading to conscious attention and adaptation of the priors that incorrectly predicted the event. However, this is not the case for repressed priors (RPs), which are defined by the inability to become reconsolidated or removed, despite ongoing error signal production. We hypothesize that Solms' RPs correspond with the conflictual complexes, as described by Moser in his dream formation theory. Thus, in dreams and dream-like states, these unconscious RPs might become accessible in symbolic and non-declarative forms that the subject is able to feel and make sense of. Finally, we present the similarities between dreaming and the psychedelic state. Insights from psychedelic research could be used to inform dream research and related therapeutic interventions, and vice versa. We propose further empirical research questions and methods and finally present our ongoing trial "Biological Functions of Dreaming" to test the hypothesis that dreaming predicts intact sleep architecture and memory consolidation, via a lesion model with stroke patients who lost the ability to dream.

Role of cerebellum in sleep-dependent memory processes

The activities and role of the cerebellum in sleep have, until recently, been largely ignored by both the sleep and cerebellum fields. Human sleep studies often neglect the cerebellum because it is at a position in the skull that is inaccessible to EEG electrodes. Animal neurophysiology sleep studies have focussed mainly on the neocortex, thalamus and the hippocampus. However, recent neurophysiological studies have shown that not only does the cerebellum participate in the sleep cycle, but it may also be implicated in off-line memory consolidation. Here we review the literature on cerebellar activity during sleep and the role it plays in off-line motor learning, and introduce a hypothesis whereby the cerebellum continues to compute internal models during sleep that train the neocortex.

What are the neural mechanisms and physiological functions of dreams?

Dreams are mental experiences, including perceptions, thoughts, and emotions, that occur during sleep. In dreams, hallucinatory perceptions, particularly visual and motoric, are often accompanied by negative emotions. When people dream, they perceive them as real even though they are bizarre and distorted in time and space. People often cannot recall their dreams, even though people dream every night. Dreaming is a strange physiological phenomenon. Research has demonstrated that dreaming is closely associated with rapid eye movement (REM) sleep. It is known that dreaming also occurs during non-REM (NREM) sleep, but the content appears to be different. Dreams during REM sleep tend to be longer, more vivid, more story-like, and more bizarre than those during NREM sleep. In this review, the neural circuits underlying dreaming and the physiological functions associated with it are summarized. Two major theories have been proposed regarding the neural circuits involved in dreaming. One is that dreams are generated by the activation of neural activity in the brainstem and its signal transmission to the cortex. The other is that dreams are caused by forebrain activation by dopamine. Whereas the physiological function of dreams remains unclear, several hypotheses have been proposed that are associated with memory and emotions.

Functional roles of REM sleep

Rapid eye movement (REM) sleep is an enigmatic and intriguing sleep state. REM sleep differs from non-REM sleep by its characteristic brain activity and from wakefulness by a reduced anti-gravity muscle tone. In addition to these key traits, diverse physiological phenomena appear across the whole body during REM sleep. However, it remains unclear whether these phenomena are the causes or the consequences of REM sleep. Experimental approaches using humans and animal models have gradually revealed the functional roles of REM sleep. Extensive efforts have been made to interpret the characteristic brain activity in the context of memory functions. Numerous physical and psychological functions of REM sleep have also been proposed. Moreover, REM sleep has been implicated in aspects of brain development. Here, we review the variety of functional roles of REM sleep, mainly as revealed by animal models. In addition, we discuss controversies regarding the functional roles of REM sleep.

No evidence for a preferential role of sleep in episodic memory abstraction

Substantial evidence suggests that sleep has a role in declarative memory consolidation. An influential notion holds that such sleep-related memory consolidation is associated with a process of abstraction. The neural underpinnings of this putative process are thought to involve a hippocampo-neocortical dialogue. Specifically, the idea is that, during sleep, the statistical contingencies across episodes are re-coded to a less hippocampus-dependent format, while at the same time losing configural information. Two previous studies from our lab, however, failed to show a preferential role of sleep in either episodic memory decontextualisation or the formation of abstract knowledge across episodic exemplars. Rather these processes occurred over sleep and wake time alike. Here, we present two experiments that replicate and extend these previous studies and exclude some alternative interpretations. The combined data show that sleep has no preferential function in this respect. Rather, hippocampus-dependent memories are generalised to an equal extent across both wake and sleep time. The one point on which sleep outperforms wake is actually the preservation of episodic detail of memories stored prior to sleep.

Global dissociation of the posterior amygdala from the rest of the brain during REM sleep

Rapid-eye-movement sleep (REMS) or paradoxical sleep is associated with intense neuronal activity, fluctuations in autonomic control, body paralysis and brain-wide hyperemia. The mechanisms and functions of these energy-demanding patterns remain elusive and a global picture of brain activation during REMS is currently missing. In the present work, we performed functional ultrasound imaging on rats over multiple coronal and sagittal brain sections during hundreds of spontaneous REMS episodes to provide the spatiotemporal dynamics of vascular activity in 259 brain regions spanning more than 2/3 of the total brain volume. We first demonstrate a dissociation between basal/midbrain and cortical structures, the first ones sustaining tonic activation during REMS while the others are activated in phasic bouts. Second, we isolated the vascular compartment in our recordings and identified arteries in the anterior part of the brain as strongly involved in the blood supply during REMS episodes. Finally, we report a peculiar activation pattern in the posterior amygdala, which is strikingly disconnected from the rest of the brain during most REMS episodes. This last finding suggests that the amygdala undergoes specific processing during REMS and may be linked to the regulation of emotions and the creation of dream content during this very state.

Memory loss at sleep onset

Every night, we pass through a transitory zone at the borderland between wakefulness and sleep, named the first stage of nonrapid eye movement sleep (N1). N1 sleep is associated with increased hippocampal activity and dream-like experiences that incorporate recent wake materials, suggesting that it may be associated with memory processing. Here, we investigated the specific contribution of N1 sleep in the processing of memory traces. Participants were asked to learn the precise locations of 48 objects on a grid and were then tested on their memory for these items before and after a 30-min rest during which participants either stayed fully awake or transitioned toward N1 or deeper (N2) sleep. We showed that memory recall was lower (10% forgetting) after a resting period, including only N1 sleep compared to N2 sleep. Furthermore, the ratio of alpha/theta power (an electroencephalography marker of the transition toward sleep) correlated negatively with the forgetting rate when taking into account all sleepers (N1 and N2 groups combined), suggesting a physiological index for memory loss that transcends sleep stages. Our findings suggest that interrupting sleep onset at N1 may alter sleep-dependent memory consolidation and promote forgetting.

Paradoxical somatodendritic decoupling supports cortical plasticity during REM sleep

Rapid eye movement (REM) sleep is associated with the consolidation of emotional memories. Yet, the underlying neocortical circuits and synaptic mechanisms remain unclear. We found that REM sleep is associated with a somatodendritic decoupling in pyramidal neurons of the prefrontal cortex. This decoupling reflects a shift of inhibitory balance between parvalbumin neuron-mediated somatic inhibition and vasoactive intestinal peptide-mediated dendritic disinhibition, mostly driven by neurons from the central medial thalamus. REM-specific optogenetic suppression of dendritic activity led to a loss of danger-versus-safety discrimination during associative learning and a lack of synaptic plasticity, whereas optogenetic release of somatic inhibition resulted in enhanced discrimination and synaptic potentiation. Somatodendritic decoupling during REM sleep promotes opposite synaptic plasticity mechanisms that optimize emotional responses to future behavioral stressors.

Learning cortical representations through perturbed and adversarial dreaming

Humans and other animals learn to extract general concepts from sensory experience without extensive teaching. This ability is thought to be facilitated by offline states like sleep where previous experiences are systemically replayed. However, the characteristic creative nature of dreams suggests that learning semantic representations may go beyond merely replaying previous experiences. We support this hypothesis by implementing a cortical architecture inspired by generative adversarial networks (GANs). Learning in our model is organized across three different global brain states mimicking wakefulness, non-rapid eye movement (NREM), and REM sleep, optimizing different, but complementary, objective functions. We train the model on standard datasets of natural images and evaluate the quality of the learned representations. Our results suggest that generating new, virtual sensory inputs via adversarial dreaming during REM sleep is essential for extracting semantic concepts, while replaying episodic memories via perturbed dreaming during NREM sleep improves the robustness of latent representations. The model provides a new computational perspective on sleep states, memory replay, and dreams, and suggests a cortical implementation of GANs.

Effects of Sleep on Language and Motor Consolidation: Evidence of Domain General and Specific Mechanisms

The current study explores the effects of time and sleep on the consolidation of a novel language learning task containing both item-specific knowledge and the extraction of grammatical regularities. We also compare consolidation effects in language and motor sequence learning tasks, to ask whether consolidation mechanisms are domain general. Young adults learned to apply plural inflections to novel words based on morphophonological rules embedded in the input, and learned to type a motor sequence using a keyboard. Participants were randomly assigned into one of two groups, practicing each task during either the morning or evening hours. Both groups were retested 12 and 24 hours post-training. Performance on frequent trained items in the language task stabilized only following sleep, consistent with a hippocampal mechanism for item-specific learning. However, regularity extraction, indicated by generalization to untrained items in the linguistic task, as well as performance on motor sequence learning, improved 24 hours post-training, irrespective of the timing of sleep. This consolidation process is consistent with a frontostriatal skill-learning mechanism, common across the language and motor domains. This conclusion is further reinforced by cross-domain correlations at the individual level between improvement across 24 hours in the motor task and in the low-frequency trained items in the linguistic task, which involve regularity extraction. Taken together, our results at the group and individual levels suggest that some aspects of consolidation are shared across the motor and language domains, and more specifically, between motor sequence learning and grammar learning.

Forgetting as a form of adaptive engram cell plasticity

One leading hypothesis suggests that memories are stored in ensembles of neurons (or 'engram cells') and that successful recall involves reactivation of these ensembles. A logical extension of this idea is that forgetting occurs when engram cells cannot be reactivated. Forms of 'natural forgetting' vary considerably in terms of their underlying mechanisms, time course and reversibility. However, we suggest that all forms of forgetting involve circuit remodelling that switches engram cells from an accessible state (where they can be reactivated by natural recall cues) to an inaccessible state (where they cannot). In many cases, forgetting rates are modulated by environmental conditions and we therefore propose that forgetting is a form of neuroplasticity that alters engram cell accessibility in a manner that is sensitive to mismatches between expectations and the environment. Moreover, we hypothesize that disease states associated with forgetting may hijack natural forgetting mechanisms, resulting in reduced engram cell accessibility and memory loss.

Associative memory model with arbitrary Hebbian length

Conversion of temporal to spatial correlations in the cortex is one of the most intriguing functions in the brain. The learning at synapses triggering the correlation conversion can take place in a wide integration window, whose influence on the correlation conversion remains elusive. Here we propose a generalized associative memory model of pattern sequences, in which pattern separations within an arbitrary Hebbian length are learned. The model can be analytically solved, and predicts that a small Hebbian length can already significantly enhance the correlation conversion, i.e., the stimulus-induced attractor can be highly correlated with a significant number of patterns in the stored sequence, thereby facilitating state transitions in the neural representation space. Moreover, an anti-Hebbian component is able to reshape the energy landscape of memories, akin to the memory regulation function during sleep. Our work thus establishes the fundamental connection between associative memory, Hebbian length, and correlation conversion in the brain.

Eigenvalue spectrum of neural networks with arbitrary Hebbian length

Associative memory is a fundamental function in the brain. Here, we generalize the standard associative memory model to include long-range Hebbian interactions at the learning stage, corresponding to a large synaptic integration window. In our model, the Hebbian length can be arbitrarily large. The spectral density of the coupling matrix is derived using the replica method, which is also shown to be consistent with the results obtained by applying the free probability method. The maximal eigenvalue is then obtained by an iterative equation, related to the paramagnetic to spin glass transition in the model. Altogether, this work establishes the connection between the associative memory with arbitrary Hebbian length and the asymptotic eigen-spectrum of the neural-coupling matrix.

Memory retrieval dynamics and storage capacity of a modular network model of association cortex with featural decomposition

The primate heteromodal cortex presents an evident functional modularity at a mesoscopic level, with physiological and anatomical evidence pointing to it as likely substrate of long-term memory. In order to investigate some of its properties, a model of multimodular autoassociator is studied. Each of the many modules represents a neocortical functional ensemble of recurrently connected neurons and operates as a Hebbian autoassociator, storing a number of local features which it can recall upon cue. The global memory patterns are made of combinations of features sparsely distributed across the modules. Intermodular connections are modelled as a finite-connectivity random graph. Any pair of features in any respective pair of modules is allowed to be involved in several memory patterns; the coarse-grained modular network dynamics is defined in such a way as to overcome the consequent ambiguity of associations. Effects of long-range homeostatic synaptic scaling on network performance are also assessed. The dynamical process of cued retrieval almost saturates a natural upper bound while producing negligible spurious activation. The extent of finite-size effects on storage capacity is quantitatively evaluated. In the limit of infinite size, the functional relationship between storage capacity and number of features per module reduces to that which other authors found by methods from equilibrium statistical mechanics, which suggests that the origin of the functional form is of a combinatorial nature. In contrast with its apparent inevitability at intramodular level, long-range synaptic scaling results to be of minor relevance to both retrieval and storage capacity, casting doubt on its existence in the neocortex. A conjecture is also posited about how statistical fluctuation of connectivity across the network may underpin spontaneous emergence of semantic hierarchies through learning.

How REM sleep shapes hypothalamic computations for feeding behavior

The electrical activity of diverse brain cells is modulated across states of vigilance, namely wakefulness, non-rapid eye movement (NREM) sleep, and rapid eye movement (REM) sleep. Enhanced activity of neuronal circuits during NREM sleep impacts on subsequent awake behaviors, yet the significance of their activation, or lack thereof, during REM sleep remains unclear. This review focuses on feeding-promoting cells in the lateral hypothalamus (LH) that express the vesicular GABA and glycine transporter (vgat) as a model to further understand the impact of REM sleep on neural encoding of goal-directed behavior. It emphasizes both spatial and temporal aspects of hypothalamic cell dynamics across awake behaviors and REM sleep, and discusses a role for REM sleep in brain plasticity underlying energy homeostasis and behavioral optimization.

Distributed associative memory network with memory refreshing loss

Despite recent progress in memory augmented neural network (MANN) research, associative memory networks with a single external memory still show limited performance on complex relational reasoning tasks. Especially the content-based addressable memory networks often fail to encode input data into rich enough representation for relational reasoning and this limits the relation modeling performance of MANN for long temporal sequence data. To address these problems, here we introduce a novel Distributed Associative Memory architecture (DAM) with Memory Refreshing Loss (MRL) which enhances the relation reasoning performance of MANN. Inspired by how the human brain works, our framework encodes data with distributed representation across multiple memory blocks and repeatedly refreshes the contents for enhanced memorization similar to the rehearsal process of the brain. For this procedure, we replace a single external memory with a set of multiple smaller associative memory blocks and update these sub-memory blocks simultaneously and independently for the distributed representation of input data. Moreover, we propose MRL which assists a task's target objective while learning relational information existing in data. MRL enables MANN to reinforce an association between input data and task objective by reproducing stochastically sampled input data from stored memory contents. With this procedure, MANN further enriches the stored representations with relational information. In experiments, we apply our approaches to Differential Neural Computer (DNC), which is one of the representative content-based addressing memory models and achieves the state-of-the-art performance on both memorization and relational reasoning tasks.

Integration of sleep homeostasis and navigation in Drosophila

During sleep, the brain undergoes dynamic and structural changes. In Drosophila, such changes have been observed in the central complex, a brain area important for sleep control and navigation. The connectivity of the central complex raises the question about how navigation, and specifically the head direction system, can operate in the face of sleep related plasticity. To address this question, we develop a model that integrates sleep homeostasis and head direction. We show that by introducing plasticity, the head direction system can function in a stable way by balancing plasticity in connected circuits that encode sleep pressure. With increasing sleep pressure, the head direction system nevertheless becomes unstable and a sleep phase with a different plasticity mechanism is introduced to reset network connectivity. The proposed integration of sleep homeostasis and head direction circuits captures features of their neural dynamics observed in flies and mice.

The Defensive Activation Theory: REM Sleep as a Mechanism to Prevent Takeover of the Visual Cortex

Regions of the brain maintain their territory with continuous activity: if activity slows or stops (e.g., because of blindness), the territory tends to be taken over by its neighbors. A surprise in recent years has been the speed of takeover, which is measurable within an hour. These findings lead us to a new hypothesis on the origin of REM sleep. We hypothesize that the circuitry underlying REM sleep serves to amplify the visual system's activity periodically throughout the night, allowing it to defend its territory against takeover from other senses. We find that measures of plasticity across 25 species of primates correlate positively with the proportion of rapid eye movement (REM) sleep. We further find that plasticity and REM sleep increase in lockstep with evolutionary recency to humans. Finally, our hypothesis is consistent with the decrease in REM sleep and parallel decrease in neuroplasticity with aging.

The overfitted brain: Dreams evolved to assist generalization

Understanding of the evolved biological function of sleep has advanced considerably in the past decade. However, no equivalent understanding of dreams has emerged. Contemporary neuroscientific theories often view dreams as epiphenomena, and many of the proposals for their biological function are contradicted by the phenomenology of dreams themselves. Now, the recent advent of deep neural networks (DNNs) has finally provided the novel conceptual framework within which to understand the evolved function of dreams. Notably, all DNNs face the issue of overfitting as they learn, which is when performance on one dataset increases but the network's performance fails to generalize (often measured by the divergence of performance on training versus testing datasets). This ubiquitous problem in DNNs is often solved by modelers via "noise injections" in the form of noisy or corrupted inputs. The goal of this paper is to argue that the brain faces a similar challenge of overfitting and that nightly dreams evolved to combat the brain's overfitting during its daily learning. That is, dreams are a biological mechanism for increasing generalizability via the creation of corrupted sensory inputs from stochastic activity across the hierarchy of neural structures. Sleep loss, specifically dream loss, leads to an overfitted brain that can still memorize and learn but fails to generalize appropriately. Herein this "overfitted brain hypothesis" is explicitly developed and then compared and contrasted with existing contemporary neuroscientific theories of dreams. Existing evidence for the hypothesis is surveyed within both neuroscience and deep learning, and a set of testable predictions is put forward that can be pursued both in vivo and in silico.

Teaching Dream Interpretation to Psychiatric Residents

Though the clinical and experimental literature suggests that discussing dreams with therapy patients can be beneficial in a variety of important ways, there appears to be a gap in educational opportunities for psychiatric residents regarding this process. To address this educational need at the authors' residency program, a class in psychodynamically oriented dream analysis was implemented, and data was collected in the form of learner surveys both before and after they took the course. The survey found that the level of importance placed on dream work, the comfort level in discussing dreams with patients, and the frequency with which dreams were discussed in sessions were all increased after taking the course. Our conclusion was that these preliminary results suggest that implementing a structured course on dream analysis may help to fill the educational gap.

Contributions of post-learning REM and NREM sleep to memory retrieval

It has become clear that sleep after learning has beneficial effects on the later retrieval of newly acquired memories. The neural mechanisms underlying these effects are becoming increasingly clear as well, particularly those of non-REM sleep. However, much is still unknown about the sleep and memory relationship: the sleep state or features of sleep physiology that associate with memory performance often vary by task or experimental design, and the nature of this variability is not entirely clear. This paper describes pertinent features of sleep physiology and provides a detailed review of the scientific literature indicating beneficial effects of post-learning sleep on memory retrieval. This paper additionally introduces a hypothesis which attributes these beneficial effects of post-learning sleep to separable processes of memory reinforcement and memory refinement whereby reinforcement supports one's ability to retrieve a given memory and refinement supports the precision of that memory retrieval in the context of competitive alternatives. It is observed that features of non-REM sleep are involved in a post-learning substantiation of memory representations that benefit memory performance; thus, memory reinforcement is primarily attributed to non-REM sleep. Memory refinement is primarily attributed to REM sleep given evidence of bidirectional synaptic plasticity in REM sleep and findings from studies of selective REM sleep deprivation.

Boltzmann Machines as Generalized Hopfield Networks: A Review of Recent Results and Outlooks

The Hopfield model and the Boltzmann machine are among the most popular examples of neural networks. The latter, widely used for classification and feature detection, is able to efficiently learn a generative model from observed data and constitutes the benchmark for statistical learning. The former, designed to mimic the retrieval phase of an artificial associative memory lays in between two paradigmatic statistical mechanics models, namely the Curie-Weiss and the Sherrington-Kirkpatrick, which are recovered as the limiting cases of, respectively, one and many stored memories. Interestingly, the Boltzmann machine and the Hopfield network, if considered to be two cognitive processes (learning and information retrieval), are nothing more than two sides of the same coin. In fact, it is possible to exactly map the one into the other. We will inspect such an equivalence retracing the most representative steps of the research in this field.