Kurzgesagt – In a Nutshell
Kurzgesagt – In a Nutshell
Embedded Files
Sources – Memory
Sources – Memory
Disclaimer: The brain is an extremely complex organ, and memory research is constantly evolving. Many of the names, labels, and processes used to describe memory depend on the scientific model or framework being applied. We aimed to explain the broad strokes of how memory works in humans as accurately as possible without being misleading. But to keep the video clear and easy to follow, we have to simplify and leave out many details.
We thank the following experts for their critical reading, feedback and corrections:
– Prof. Wilma Bainbridge
University of Chicago, USA
– Prof. Sheena Josselyn
University of Toronto, Canada
– Prof. Lukas Kunz
University of Bonn, Germany
– Prof. Michel van den Oever
Vrije Universiteit Amsterdam, Netherlands
– But in a nutshell, what makes you think and feel is a complex system of about 86 billion neurons.
There are different estimates regarding the total number of neurons in the human brain, and the sample sizes for the most reliable estimates, i.e. the number of complete brains analyzed at this level of detail, are relatively small. Most commonly used in the neuroscientific literature is an estimate of roughly 86 billion neurons per human brain on average, which we also chose to use here. However, there are differences between individuals, and future studies may reveal more accurate numbers. Below, we cite a 2025 review summarizing the current state of the literature.
#Goriely A. Eighty-six billion and counting: do we know the number of neurons in the human brain? Brain. 2025
https://academic.oup.com/brain/article/148/3/689/7909879
Quote: “
Experiments have shown variations between 62 and 94 bn neurons in the human brain (n = 9).
An experimental study on the number of neurons suggests an average between 73 and 99 bn neurons in the healthy male human brain (n = 4).
An experimental study on the number of neurons suggests an average between 61 and 73 bn neurons in the healthy female human brain (n = 5).
Clearly, none of these statements is satisfactory or as catchy as ‘the human brain has 86 billion neurons’. Yet, they are the true reflection of our knowledge. We cannot present a more precise assertion without more available data.”
– Extremely complex electrochemical root systems, sending and receiving signals through synapses – tiny gaps between cells, where an electrical signal is converted into chemicals, bridges the distance, is received and converted into electricity again. This is how neurons talk to each other.
Here we are describing how chemical synapses work. There are also electrical synapses that don’t use “chemicals” (neurotransmitters) to bridge the gap between cells, but instead pass on electricity directly to the adjacent neuron. Electrical synapses are even faster than chemical synapses.
#J.S. Dittman, Taking a closer look at the synapse, Proc. Natl. Acad. Sci. U.S.A. (2024)
https://doi.org/10.1073/pnas.2412457121
Quote: “Communication between neurons relies on highly specialized structures (synapses) where an activated presynaptic neuron releases packets of neurotransmitter molecules that diffuse across a narrow gap (synaptic cleft) and bind to ligand-gated ion channels (and other types of receptors) on the postsynaptic neuron. Postsynaptic receptor activation delivers the synaptic message in the form of electrical currents and intracellular signaling cascades. The extraordinary speed and functional diversity of this elaborate process have fascinated and baffled researchers for decades, while the small size of synapses (significantly less than a micron in many cases) has posed a formidable challenge for visualizing and studying the underlying synaptic machinery.”
#Heitler WJ. Neurosim 5 – Synapses. University of St. Andrews (2019)
https://www.st-andrews.ac.uk/~wjh/neurosim/TutorialV5_6/Synapses.html
Quote: “Chemical Synapses
At a chemical synapse, information is transmitted between neurons in the form of a chemical signal. Specifically, depolarization in the output region of a pre-synaptic neurone opens voltage-dependent calcium channels, and the resulting inflow of calcium triggers the exocytosis of vesicles containing the chemical signal, the neurotransmitter. This diffuses across the synaptic cleft and interacts with receptors on the post-synaptic membrane. It is this interaction which causes electrical events in the post-synaptic neuron.
In a spiking chemical synapses only a spike in the pre-synaptic neuron generates enough depolarization to cause transmitter release. In a non-spiking chemical synapses, sub-threshold pre-synaptic depolarization can cause transmitter release on its own, without the occurrence of a spike.
[...]
Electrical Synapses
Electrical synapses are mediated by gap junctions that form an electrically-conducting cytoplasmic bridge from one neuron to another. This means that current can flow intracellularly between neurons, so that if one neuron spikes, the depolarizing current entering that neuron spreads to the other neuron, which consequently receives an electrical EPSP. This has two important consequences. First, electrical synapses are faster than chemical synapses because they do not involve the intermediary chemical signal. This means that electrical synapses are often found in neural circuits mediating behaviours where speed is essential, such as escape behaviours. Second, most electrical synapses can “share” both excitation and inhibition between pools of neurons, thus tending to synchronize them. So electrical synapses are often found in pools of neurons that perform similar functions and which often spike together, such as the motor unit pools innervating a particular muscle.”
#Towlson E, Vértes P, Yan G, Chew YL, Walker D, Schafer W & Barabasi, A-L. (2018). Caenorhabditis elegans and the network control framework - FAQs.
“Figure 2. Chemical synapses and electrical gap junctions. Chemical synapses and electrical gap junctions have very different properties and underlying mechanisms. In electrical gap junctions (left), voltage is transferred via touching membranes and signals may pass in both directions. In chemical synapses (right), signals are transferred through ion channels from the pre-to post-synaptic neuron.“
– And it's a busy conversation because a typical neuron is connected with up to 10,000 others.
The number of connections with other neurons (synapses) varies strongly depending on the type of neuron and also within types. For the human brain, an “average neuron” is often quoted as having thousands of synapses, but detailed measurements in the cortex indicate that some types of neurons have tens of thousands of synapses each.
#Nowinski WL. On human nanoscale synaptome: Morphology modeling and storage estimation. PLOS ONE. (2024)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0310156
Quote: “The number of synapses per neuron is also highly variable. Kandel et al. estimate that there are around 1014 (100 trillion) synapses in the average adult human brain [18] (meaning 1,000 synapses per neuron). Ref. [38] reports that the total number of synapses in the human neocortex is approximately 0.15×1015 (0.15 quadrillion). According to a reference book [39], the average neuron has 1,000 synapses with other neurons. Ten thousand connections per neuron are reported in [23]. Some sources estimate that a single neuron can have between 1,000–15,000 synaptic connections [24,40]. Defelipe demonstrated that the number of synapses per neuron is also cortical layer-dependent and the study of the synaptic profiles within cubes of cortical tissue of 50 μm wide by 50 μm thick revealed that for 158 neurons and 4,483,400 synapses examined in 6 cortical layers, the average number of synapses per neuron was 29,642 (100,042 in layer I (for 5 neurons only), 17,046 in layer II, 37,066 in layer IIIa, 56,521 in layer IIIb, 15,989 in layer IV, 29,965 in layer V, and 28,224 in layer VI) [41].”
– Together they’re a network of hundreds of trillions of connections.
It is estimated that there are roughly 100–600 trillion synapses in the adult human brain, perhaps even more.
#Johansen A, Beliveau V, Colliander E, Raval NR, Dam VH, Gillings N, Aznar S, Svarer C, Plavén-Sigray P, Knudsen GM. An In Vivo High-Resolution Human Brain Atlas of Synaptic Density. J Neurosci. 2024
https://www.jneurosci.org/content/44/33/e1750232024
Quote: “Synapses are the junctions between neurons and constitute a central site for neuronal signaling in the brain. Estimates of the number of synapses formed on a single neuron in the adult human brain vary greatly (7,200–80,000), bringing their total number to a staggering range of 6 × 1014–7 × 1015 (Silbereis et al., 2016).”
#Nowinski WL. On human nanoscale synaptome: Morphology modeling and storage estimation. PLOS ONE. (2024)
https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0310156
Quote: “The number of synapses per neuron is also highly variable. Kandel et al. estimate that there are around 1014 (100 trillion) synapses in the average adult human brain [18] (meaning 1,000 synapses per neuron). Ref. [38] reports that the total number of synapses in the human neocortex is approximately 0.15×1015 (0.15 quadrillion).”
– A hundred times more than galaxies in the observable universe.
The observable universe is estimated to contain roughly 2 trillion galaxies. Versus the roughly 100-600 trillion neurons in the human brain, that is about a 50 to 300-fold difference. For our visuals, we compare a conservative 200 trillion synapses to 2 trillion galaxies, which is a 100-fold difference.
#Conselice, C. J., Wilkinson, A., Duncan, K., & Mortlock, A. The evolution of galaxy number density at z < 8 and its implications. Astrophysical Journal. (2016)
https://doi.org/10.3847/0004-637X/830/2/83
Quote: “The evolution of the number density of galaxies in the universe, and thus also the total number of galaxies, is a fundamental question with implications for a host of astrophysical problems including galaxy evolution and cosmology. However, there has never been a detailed study of this important measurement, nor a clear path to answer it. To address this we use observed galaxy stellar mass functions up to z ∼ 8 to determine how the number densities of galaxies change as a function of time and mass limit. We show that the increase in the total number density of galaxies (ϕT), more massive than M* = 106 M⊙, decreases as ϕT ∼ t−1, where t is the age of the universe. We further show that this evolution turns over and rather increases with time at higher mass lower limits of M* > 107 M⊙. By using the M* = 106 M⊙ lower limit we further show that the total number of galaxies in the universe up to z = 8 is
(2 trillion), almost a factor of 10 higher than would be seen in an all sky survey at Hubble Ultra-Deep Field depth. We discuss the implications for these results for galaxy evolution, as well as compare our results with the latest models of galaxy formation. These results also reveal that the cosmic background light in the optical and near-infrared likely arise from these unobserved faint galaxies. We also show how these results solve the question of why the sky at night is dark, otherwise known as Olbers’ paradox.”
– From these connections the magic of your existence emerges. To create purpose in this chaos some connections need to grow stronger. Whenever two neurons fire together their synapses change and their connection gets stronger. They become buddies if you want, and when their buddy calls they join in.
Synaptic plasticity is the ability of synapses to strengthen or weaken over time, in response to increases or decreases in their activity. “Hebbian plasticity” is a theory describing a type of synaptic plasticity, where the synaptic connections strengthen when a presynaptic neuron repeatedly helps fire a postsynaptic neuron.
The theory is often summarized as "Neurons that fire together, wire together."
#Friedemann Zenke, Wulfram Gerstner; Hebbian plasticity requires compensatory processes on multiple timescales. Philos Trans R Soc Lond B Biol Sci. 2017
https://doi.org/10.1098/rstb.2016.0259
Quote: “Hebbian plasticity is a form of synaptic plasticity which is induced by and further amplifies correlations in neuronal activity. It has been observed in many brain areas and can be induced quickly on a timescale of seconds to minutes. Its effect, however, is often long-lasting. It can last hours, days and possibly a lifetime. Owing to these properties, Hebbian plasticity is widely assumed to be the neural basis of associative long-term memory [2–4]. Moreover, Hebbian learning is thought to be the basis of developmental changes such as receptive field development [5–9].”
– If we zoom out, patterns emerge. Dozens or thousands of neurons organize into local columns.
Columns are small, vertical anatomical structures in the neocortex region of the human brain. Similar structures also exist in other animals, for example they are very well-studied in rodents. Their size varies widely: the number of neurons per column can range from hundreds (so-called “minicolumns”) to tens of thousands of cells. Of note, there is no singular definition of what constitutes a cortical column (in terms of its size/borders and function), so the estimates for how many neurons they contain vary accordingly.
#Barbas H, Zikopoulos B and John YJ. The inevitable inequality of cortical columns. Front. Syst. Neurosci. (2022)
https://www.frontiersin.org/journals/systems-neuroscience/articles/10.3389/fnsys.2022.921468/full
Quote: “The concept of cortical column emerged over 50 years ago from findings in the somatosensory system (reviewed in Mountcastle, 1997). In its initial description, the term captured Mountcastle’s observation that neurons recorded along vertical penetrations from the top cellular layer 2 to layer 6 responded to the same stimuli, such as light touch on the body surface (reviewed in Kaas, 2012). Mountcastle’s work was followed by findings from the primary visual cortex of cats and monkeys by Hubel and Wiesel (e.g., Hubel and Wiesel, 1968). The latter described vertical columns of neurons that preferentially respond to stimuli of a specific orientation. Columns of best frequency responses were also mapped on the primary auditory cortex of macaques (e.g., Merzenich and Brugge, 1973). The organization within the primary motor cortex was more complex, whereby evoked movement of a joint was clustered in “mini-columns,” while adjoining mini columns above or below evoked responses to a different joint, likely associated with the complex sequences required for responses in the cortical motor system (reviewed in Kaas, 2012).
The column concept thus began as a functional principle. The introduction of neural tracers to study connections revealed patterns that could be the anatomic correlates of physiologic columns, as seen widely throughout the cortex, including high-order association prefrontal areas (e.g., Bugbee and Goldman-Rakic, 1983; Rockland, 2010; Kaas, 2012; Casanova and Casanova, 2019). Connection studies and molecular markers sparked new debates as labeled patterns often were modular, resembling short columns. The cytochrome oxidase marker, for example, labeled patches (modules), found especially in layers 3 and 2 of primary visual cortex (V1), where neurons did not respond to a specific orientation of a visual stimulus, as neurons did above and below the blobs (Livingstone and Hubel, 1984). Debates about the features of columns or modules, such as extension of axons beyond their borders could be reconciled by the presence of inhibitory neurons, which can reduce extraneous responses on the flanks of active columns, attesting to their dynamic nature (da Costa and Martin, 2010; Rockland, 2010; Wang, 2020).
Neurons in columns or modules are strongly interconnected locally in the vertical direction, and are also innervated by fewer but highly consequential extrinsic connections from other cortices or from subcortical structures (Gilbert, 1983; Pandya et al., 1988; Callaway, 1998). Discussions about the organization of columns/modules have mostly proceeded outside such considerations, outside the context of the general principle of cortical systematic variation, and outside the intricate confluence of the vertical (columnar) and horizontal (laminar) organization of the cortex, which we address here.”
#Rockland, K., & Molnár, Z. The Cortical Column. Oxford Research Encyclopedia of Neuroscience (2025)
Quote: “The term “cortical column” suggests some degree of verticality, perpendicular to the pia surface, but it is a surprisingly ambiguous term. An important clarification is that there are multiple different anatomical structures that qualify as “columnar,” but at different scales. Small-scale minicolumns (~50–100 μm in diameter) include cellular cords (also called strings or stacks), bundles of apical dendrites and myelinated axons, and vertically aligned axons of double bouquet cells. Larger scale macrocolumns (“modules,” 250–500 μm), include extrinsic and intrinsic cortical connections. In contrast with the implication of “through depth columnar” organization, none of these extend from layer 1 to white matter. For all of these, there are multiple area- and species-specific variations. A major challenge is how the multi-scale elements of brain organization (layers and columns, minicolumns and macrocolumns) cooperate and work together to translate and interface with the multiple disparate brain functions simultaneously ongoing at any one time.”
#Onesto, V., Villani, M., Narducci, R. et al. Cortical-like mini-columns of neuronal cells on zinc oxide nanowire surfaces. Sci Rep (2019)
https://www.nature.com/articles/s41598-019-40548-z
Quote: “Notably, different works estimated the number of neurons forming a single minicolumn in the primate cortex, finding a constant range spanning from 80 to 11082,85 (but see the striate cortex with a value 2.5 times larger82). Interestingly, the number of neurons per column resulted constant across five diverse species from different mammalian orders85.”
#S. Herculano-Houzel, C.E. Collins, P. Wong, J.H. Kaas, & R. Lent, The basic nonuniformity of the cerebral cortex, Proc. Natl. Acad. Sci. U.S.A. (2008)
https://www.pnas.org/doi/full/10.1073/pnas.0805417105
Quote: “One study across several mammalian orders found that the number of neurons beneath 1 mm2 of cerebral cortex varies from 20,000 to >160,000 (14). A more recent study found that the number of neurons beneath 1 mm2 of the cerebral cortex decreases from >50,000 to <20,000 across delphinid species with increasing brain size (12)—and, therefore, probably decreases with decreasing neuronal density.”
– These columns are very basic information processing units that process a tiny piece of all the input coming in from your senses: like light or dark, a location in space, how a texture feels, the sound of words and so on. You have columns for sound, images, touch etc, all located in different parts of your brain.
Sensory perceptions are processed in different parts of the brain depending on the sense. Columns and microcolumns – vertically organized groups of neurons – play a central role in sensory processing. They occur throughout the cerebral cortex, including the brain regions relevant for sensory processing, and act as modular microcircuits that extract, transform, and forward relevant signals.
Of note, other parts of the brain that also play a big role in memory formation, such as e.g. the hippocampus, do not exhibit columnar structures in their neuronal firing patterns. But here we are focussing on processing sensory input, and columns are very important for that.
#Kashyap, S., Ivanov, D., Havlicek, M. et al. Resolving laminar activation in human V1 using ultra-high spatial resolution fMRI at 7T. Sci Rep (2018). https://doi.org/10.1038/s41598-018-35333-3
Quote: “The human cortex is structured into a myriad of structural and functional units1. These units can be indexed along both the normal and tangential coordinates with respect to the surface of the cortex2. The normal coordinates represent the cortical depth, from the pial to the white-matter boundary (WMB), also called cortical layers or laminae. The tangential coordinates represent the distance along the cortical ribbon, and the structural-functional units along this axis are called brain areas and columns, on the macroscopic and mesoscopic scales, respectively. These subdivisions are characterized by their distinctive histological profiles3,4 or by their specific responsiveness to external stimuli and role in cognitive and physiological processes5,6.”
#Haueis P. The life of the cortical column: opening the domain of functional architecture of the cortex (1955-1981). Hist Philos Life Sci. 2016
https://pmc.ncbi.nlm.nih.gov/articles/PMC4914527/
Quote: “The cortical column designates vertical structures that span all layers of the cerebral cortex, and in which neurons show similar responses to sensory stimuli. The original definition of the column goes back to Vernon Mountcastle’s mapping of single-cell recordings in the somatic sensory cortex of the cat (Mountcastle et al. 1955). It has subsequently become a basic concept that has guided over 50 years of neuroscientific research (Shepherd 2010), in which fundamental questions about the modularity of the cortex and basic principles of sensory information processing were empirically investigated. After physiologists successfully applied the column concept to many sensory and some “higher” areas of brain function, it came to be viewed as an “elementary unit of organization” of the entire cerebral cortex (Mountcastle 1957, p. 430; see also 1978, p. 16; 1997, p. 701).”
#Jia K, Goebel R, Kourtzi Z. Ultra-High Field Imaging of Human Visual Cognition. Annu Rev Vis Sci. 2023
https://www.annualreviews.org/content/journals/10.1146/annurev-vision-111022-123830
Quote: “Functional magnetic resonance imaging (fMRI), the key methodology for mapping the functions of the human brain in a noninvasive manner, is limited by low temporal and spatial resolution. Recent advances in ultra-high field (UHF) fMRI provide a mesoscopic (i.e., submillimeter resolution) tool that allows us to probe laminar and columnar circuits, distinguish bottom-up versus top-down pathways, and map small subcortical areas. We review recent work demonstrating that UHF fMRI provides a robust methodology for imaging the brain across cortical depths and columns that provides insights into the brain's organization and functions at unprecedented spatial resolution, advancing our understanding of the fine-scale computations and interareal communication that support visual cognition.
[...]
Recent advances in ultra-high field (UHF) imaging techniques provide a mesoscopic tool ( Figure 1 ) that enables us to interrogate brain computations at a finer scale (i.e., submillimeter resolution) than that afforded by standard 3T fMRI (Goense et al. 2016). In particular, UHF fMRI allows us to probe cortical layers that are known to have distinct connectivity with upstream and downstream areas, allowing us to discern bottom-up versus top-down pathways and infer the direction of information flow across areas. Furthermore, UHF fMRI enables us to map small subcortical areas [e.g., the lateral geniculate nucleus (LGN), habenula] and columnar organization [e.g., ocular dominance columns (ODCs), orientation columns], allowing us to characterize fine-scale activity that supports intra-areal computations.”
#Ai H, Lin W, Liu C, Chen N, Zhang P. Mesoscale functional organization and connectivity of color, disparity, and naturalistic texture in human second visual area. Elife. 2025
https://pmc.ncbi.nlm.nih.gov/articles/PMC11925451/
Quote: “Although parallel processing has been extensively studied in the low-level geniculostriate pathway and the high-level dorsal and ventral visual streams, less is known at the intermediate-level visual areas. In this study, we employed high-resolution fMRI at 7T to investigate the columnar and laminar organizations for color, disparity, and naturalistic texture in the human secondary visual cortex (V2), and their informational connectivity with lower- and higher-order visual areas. Although fMRI activations in V2 showed reproducible interdigitated color-selective thin and disparity-selective thick ‘stripe’ columns, we found no clear evidence of columnar organization for naturalistic textures. Cortical depth-dependent analyses revealed the strongest color-selectivity in the superficial layers of V2, along with both feedforward and feedback informational connectivity with V1 and V4. Disparity selectivity was similar across different cortical depths of V2, which showed significant feedforward and feedback connectivity with V1 and V3ab. Interestingly, the selectivity for naturalistic texture was strongest in the deep layers of V2, with significant feedback connectivity from V4. Thus, while local circuitry within cortical columns is crucial for processing color and disparity information, feedback signals from V4 are involved in generating the selectivity for naturalistic textures in area V2.”
#Moerel M, De Martino F, Uğurbil K, Formisano E, Yacoub E. Evaluating the Columnar Stability of Acoustic Processing in the Human Auditory Cortex. J Neurosci. 2018
https://pmc.ncbi.nlm.nih.gov/articles/PMC6125808/
Quote: “Using ultra-high field fMRI, we explored the cortical depth-dependent stability of acoustic feature preference in human auditory cortex. We collected responses from human auditory cortex (subjects from either sex) to a large number of natural sounds at submillimeter spatial resolution, and observed that these responses were well explained by a model that assumes neuronal population tuning to frequency-specific spectrotemporal modulations. We observed a relatively stable (columnar) tuning to frequency and temporal modulations. However, spectral modulation tuning was variable throughout the cortical depth. This difference in columnar stability between feature maps could not be explained by a difference in map smoothness, as the preference along the cortical sheet varied in a similar manner for the different feature maps. Furthermore, tuning to all three features was more columnar in primary than nonprimary auditory cortex. The observed overall lack of overlapping columnar regions across acoustic feature maps suggests, especially for primary auditory cortex, a coding strategy in which across cortical depths tuning to some features is kept stable, whereas tuning to other features systematically varies.”
#F. De Martino, M. Moerel, K. Ugurbil, R. Goebel, E. Yacoub, & E. Formisano, Frequency preference and attention effects across cortical depths in the human primary auditory cortex, Proc. Natl. Acad. Sci. U.S.A. (2015)
https://doi.org/10.1073/pnas.1507552112
Quote: “Columnar arrangements of neurons with similar preference have been suggested as the fundamental processing units of the cerebral cortex. Within these columnar arrangements, feed-forward information enters at middle cortical layers whereas feedback information arrives at superficial and deep layers. This interplay of feed-forward and feedback processing is at the core of perception and behavior. Here we provide in vivo evidence consistent with a columnar organization of the processing of sound frequency in the human auditory cortex. We measure submillimeter functional responses to sound frequency sweeps at high magnetic fields (7 tesla) and show that frequency preference is stable through cortical depth in primary auditory cortex. Furthermore, we demonstrate that—in this highly columnar cortex—task demands sharpen the frequency tuning in superficial cortical layers more than in middle or deep layers. These findings are pivotal to understanding mechanisms of neural information processing and flow during the active perception of sounds.”
#Kalyani A, Contier O, Klemm L, Azañon E, Schreiber S, Speck O, Reichert C, Kuehn E. Reduced dimension stimulus decoding and column-based modeling reveal architectural differences of primary somatosensory finger maps between younger and older adults. Neuroimage. 2023
https://www.sciencedirect.com/science/article/pii/S1053811923005815?via%3Dihub
Quote: “The primary somatosensory cortex (SI) contains fine-grained tactile representations of the body, arranged in an orderly fashion. The use of ultra-high resolution fMRI data to detect group differences, for example between younger and older adults’ SI maps, is challenging, because group alignment often does not preserve the high spatial detail of the data. Here, we use robust-shared response modeling (rSRM) that allows group analyses by mapping individual stimulus-driven responses to a lower dimensional shared feature space, to detect age-related differences in tactile representations between younger and older adults using 7T-fMRI data. Using this method, we show that finger representations are more precise in Brodmann-Area (BA) 3b and BA1 compared to BA2 and motor areas, and that this hierarchical processing is preserved across age groups. By combining rSRM with column-based decoding (C-SRM), we further show that the number of columns that optimally describes finger maps in SI is higher in younger compared to older adults in BA1, indicating a greater columnar size in older adults’ SI. Taken together, we conclude that rSRM is suitable for finding fine-grained group differences in ultra-high resolution fMRI data, and we provide first evidence that the columnar architecture in SI changes with increasing age.”
– These columns are the gears of the biological machine that is your cortex. It’s the fundamental hardware you emerge from. Everything you see, hear, or feel causes the gears to move – which means your neurons to fire together.
Processes in the brain are extremely complex, and we simplify a lot here. But in a nutshell, cortical columns are repeating little circuits that each interpret a small part of a sensory signal, and together they build what you perceive. When you perceive something, specific columns in parts of your brain fire in highly complex patterns: higher or lower firing rates, synchronous or asynchronous firing, waves of activity spreading to nearby columns etc. Together, these firing patterns serve to pre-process the perceived image, sound or other sensory input.
The firing pattern also depends on the type of stimulus: visual, audio and somatosensory are thought to be organized in a columnar structure during processing in the brain. Other senses (e.g. taste, smell) are less well studied and may also be processed in a different way, e.g. more distributed around the brain and less so in a columnar structure.
Brewer AA and Barton B. Cortical field maps across human sensory cortex. Front. Comput. Neurosci. (2023)
Quote: “Topographical representations of sensory information are emerging as a fundamental organizational pattern for perceptual processing across sensory cortex in numerous mammalian species (Kaas, 1997; Wandell et al., 2005; Krubitzer, 2007; Sanchez-Panchuelo et al., 2010; Barton et al., 2012; Prinster et al., 2017; Yushu Chen et al., 2021). Organized topographies within sensory pathways are thought to support the comparison and combination of the information carried by the various specialized neuronal populations. To enhance the brain’s ability to discriminate among different stimuli, sensory neurons that respond to similar features are frequently organized into distinct clusters or columns, and their response characteristics exhibit smooth transitions across the cortical surface. The orderly connectivity arising from such organization is likely important for increasing the efficiency of such local processes as lateral inhibition and gain control and may provide a framework for sensory processing across the sensory hierarchy (Mitchison, 1991; Van Essen, 2003; Chklovskii and Koulakov, 2004; Shapley et al., 2007; Moradi and Heeger, 2009).”
#F. De Martino, M. Moerel, K. Ugurbil, R. Goebel, E. Yacoub, & E. Formisano, Frequency preference and attention effects across cortical depths in the human primary auditory cortex, Proc. Natl. Acad. Sci. U.S.A. (2015)
https://doi.org/10.1073/pnas.1507552112
Quote: “Columnar arrangements of neurons with similar preference have been suggested as the fundamental processing units of the cerebral cortex. Within these columnar arrangements, feed-forward information enters at middle cortical layers whereas feedback information arrives at superficial and deep layers. This interplay of feed-forward and feedback processing is at the core of perception and behavior. Here we provide in vivo evidence consistent with a columnar organization of the processing of sound frequency in the human auditory cortex. We measure submillimeter functional responses to sound frequency sweeps at high magnetic fields (7 tesla) and show that frequency preference is stable through cortical depth in primary auditory cortex. Furthermore, we demonstrate that—in this highly columnar cortex—task demands sharpen the frequency tuning in superficial cortical layers more than in middle or deep layers. These findings are pivotal to understanding mechanisms of neural information processing and flow during the active perception of sounds.”
#Lübke J, Feldmeyer D. Excitatory signal flow and connectivity in a cortical column: focus on barrel cortex. Brain Struct Funct. 2007
Excitatory signal flow and connectivity in a cortical column: focus on barrel cortex - PubMed
Quote: “A basic feature of the neocortex is its organization in functional, vertically oriented columns, recurring modules of signal processing and a system of transcolumnar long-range horizontal connections. These columns, together with their network of neurons, present in all sensory cortices, are the cellular substrate for sensory perception in the brain. Cortical columns contain thousands of neurons and span all cortical layers. They receive input from other cortical areas and subcortical brain regions and in turn their neurons provide output to various areas of the brain. The modular concept presumes that the neuronal network in a cortical column performs basic signal transformations, which are then integrated with the activity in other networks and more extended brain areas. To understand how sensory signals from the periphery are transformed into electrical activity in the neocortex it is essential to elucidate the spatial-temporal dynamics of cortical signal processing and the underlying neuronal 'microcircuits'. In the last decade the 'barrel' field in the rodent somatosensory cortex, which processes sensory information arriving from the mysticial vibrissae, has become a quite attractive model system because here the columnar structure is clearly visible. In the neocortex and in particular the barrel cortex, numerous neuronal connections within or between cortical layers have been studied both at the functional and structural level. Besides similarities, clear differences with respect to both physiology and morphology of synaptic transmission and connectivity were found. It is therefore necessary to investigate each neuronal connection individually, in order to develop a realistic model of neuronal connectivity and organization of a cortical column. This review attempts to summarize recent advances in the study of individual microcircuits and their functional relevance within the framework of a cortical column, with emphasis on excitatory signal flow.”
#Jia K, Goebel R, Kourtzi Z. Ultra-High Field Imaging of Human Visual Cognition. Annu Rev Vis Sci. 2023
https://www.annualreviews.org/content/journals/10.1146/annurev-vision-111022-123830
Quote: “Functional magnetic resonance imaging (fMRI), the key methodology for mapping the functions of the human brain in a noninvasive manner, is limited by low temporal and spatial resolution. Recent advances in ultra-high field (UHF) fMRI provide a mesoscopic (i.e., submillimeter resolution) tool that allows us to probe laminar and columnar circuits, distinguish bottom-up versus top-down pathways, and map small subcortical areas. We review recent work demonstrating that UHF fMRI provides a robust methodology for imaging the brain across cortical depths and columns that provides insights into the brain's organization and functions at unprecedented spatial resolution, advancing our understanding of the fine-scale computations and interareal communication that support visual cognition.
[...]
Recent advances in ultra-high field (UHF) imaging techniques provide a mesoscopic tool ( Figure 1 ) that enables us to interrogate brain computations at a finer scale (i.e., submillimeter resolution) than that afforded by standard 3T fMRI (Goense et al. 2016). In particular, UHF fMRI allows us to probe cortical layers that are known to have distinct connectivity with upstream and downstream areas, allowing us to discern bottom-up versus top-down pathways and infer the direction of information flow across areas. Furthermore, UHF fMRI enables us to map small subcortical areas [e.g., the lateral geniculate nucleus (LGN), habenula] and columnar organization [e.g., ocular dominance columns (ODCs), orientation columns], allowing us to characterize fine-scale activity that supports intra-areal computations.”
#Ai H, Lin W, Liu C, Chen N, Zhang P. Mesoscale functional organization and connectivity of color, disparity, and naturalistic texture in human second visual area. Elife. 2025
https://pmc.ncbi.nlm.nih.gov/articles/PMC11925451/
Quote: “Although parallel processing has been extensively studied in the low-level geniculostriate pathway and the high-level dorsal and ventral visual streams, less is known at the intermediate-level visual areas. In this study, we employed high-resolution fMRI at 7T to investigate the columnar and laminar organizations for color, disparity, and naturalistic texture in the human secondary visual cortex (V2), and their informational connectivity with lower- and higher-order visual areas. Although fMRI activations in V2 showed reproducible interdigitated color-selective thin and disparity-selective thick ‘stripe’ columns, we found no clear evidence of columnar organization for naturalistic textures. Cortical depth-dependent analyses revealed the strongest color-selectivity in the superficial layers of V2, along with both feedforward and feedback informational connectivity with V1 and V4. Disparity selectivity was similar across different cortical depths of V2, which showed significant feedforward and feedback connectivity with V1 and V3ab. Interestingly, the selectivity for naturalistic texture was strongest in the deep layers of V2, with significant feedback connectivity from V4. Thus, while local circuitry within cortical columns is crucial for processing color and disparity information, feedback signals from V4 are involved in generating the selectivity for naturalistic textures in area V2.”
#Moerel M, De Martino F, Uğurbil K, Formisano E, Yacoub E. Evaluating the Columnar Stability of Acoustic Processing in the Human Auditory Cortex. J Neurosci. 2018
https://pmc.ncbi.nlm.nih.gov/articles/PMC6125808/
Quote: “Using ultra-high field fMRI, we explored the cortical depth-dependent stability of acoustic feature preference in human auditory cortex. We collected responses from human auditory cortex (subjects from either sex) to a large number of natural sounds at submillimeter spatial resolution, and observed that these responses were well explained by a model that assumes neuronal population tuning to frequency-specific spectrotemporal modulations. We observed a relatively stable (columnar) tuning to frequency and temporal modulations. However, spectral modulation tuning was variable throughout the cortical depth. This difference in columnar stability between feature maps could not be explained by a difference in map smoothness, as the preference along the cortical sheet varied in a similar manner for the different feature maps. Furthermore, tuning to all three features was more columnar in primary than nonprimary auditory cortex. The observed overall lack of overlapping columnar regions across acoustic feature maps suggests, especially for primary auditory cortex, a coding strategy in which across cortical depths tuning to some features is kept stable, whereas tuning to other features systematically varies.”
#F. De Martino, M. Moerel, K. Ugurbil, R. Goebel, E. Yacoub, & E. Formisano, Frequency preference and attention effects across cortical depths in the human primary auditory cortex, Proc. Natl. Acad. Sci. U.S.A. (2015)
https://doi.org/10.1073/pnas.1507552112
Quote: “Columnar arrangements of neurons with similar preference have been suggested as the fundamental processing units of the cerebral cortex. Within these columnar arrangements, feed-forward information enters at middle cortical layers whereas feedback information arrives at superficial and deep layers. This interplay of feed-forward and feedback processing is at the core of perception and behavior. Here we provide in vivo evidence consistent with a columnar organization of the processing of sound frequency in the human auditory cortex. We measure submillimeter functional responses to sound frequency sweeps at high magnetic fields (7 tesla) and show that frequency preference is stable through cortical depth in primary auditory cortex. Furthermore, we demonstrate that—in this highly columnar cortex—task demands sharpen the frequency tuning in superficial cortical layers more than in middle or deep layers. These findings are pivotal to understanding mechanisms of neural information processing and flow during the active perception of sounds.”
#Kalyani A, Contier O, Klemm L, Azañon E, Schreiber S, Speck O, Reichert C, Kuehn E. Reduced dimension stimulus decoding and column-based modeling reveal architectural differences of primary somatosensory finger maps between younger and older adults. Neuroimage. 2023
https://www.sciencedirect.com/science/article/pii/S1053811923005815?via%3Dihub
Quote: “The primary somatosensory cortex (SI) contains fine-grained tactile representations of the body, arranged in an orderly fashion. The use of ultra-high resolution fMRI data to detect group differences, for example between younger and older adults’ SI maps, is challenging, because group alignment often does not preserve the high spatial detail of the data. Here, we use robust-shared response modeling (rSRM) that allows group analyses by mapping individual stimulus-driven responses to a lower dimensional shared feature space, to detect age-related differences in tactile representations between younger and older adults using 7T-fMRI data. Using this method, we show that finger representations are more precise in Brodmann-Area (BA) 3b and BA1 compared to BA2 and motor areas, and that this hierarchical processing is preserved across age groups. By combining rSRM with column-based decoding (C-SRM), we further show that the number of columns that optimally describes finger maps in SI is higher in younger compared to older adults in BA1, indicating a greater columnar size in older adults’ SI. Taken together, we conclude that rSRM is suitable for finding fine-grained group differences in ultra-high resolution fMRI data, and we provide first evidence that the columnar architecture in SI changes with increasing age.”
#Avery JA, Liu AG, Ingeholm JE, Riddell CD, Gotts SJ, Martin A. Taste Quality Representation in the Human Brain. J Neurosci. 2020
https://pmc.ncbi.nlm.nih.gov/articles/PMC6989007/
Quote: “In the mammalian brain, the insula is the primary cortical substrate involved in the perception of taste. Recent imaging studies in rodents have identified a “gustotopic” organization in the insula, whereby distinct insula regions are selectively responsive to one of the five basic tastes. However, numerous studies in monkeys have reported that gustatory cortical neurons are broadly-tuned to multiple tastes, and tastes are not represented in discrete spatial locations. Neuroimaging studies in humans have thus far been unable to discern between these two models, though this may be because of the relatively low spatial resolution used in taste studies to date. In the present study, we examined the spatial representation of taste within the human brain using ultra-high resolution functional magnetic resonance imaging (MRI) at high magnetic field strength (7-tesla). During scanning, male and female participants tasted sweet, salty, sour, and tasteless liquids, delivered via a custom-built MRI-compatible tastant-delivery system. Our univariate analyses revealed that all tastes (vs tasteless) activated primary taste cortex within the bilateral dorsal mid-insula, but no brain region exhibited a consistent preference for any individual taste. However, our multivariate searchlight analyses were able to reliably decode the identity of distinct tastes within those mid-insula regions, as well as brain regions involved in affect and reward, such as the striatum, orbitofrontal cortex, and amygdala. These results suggest that taste quality is not represented topographically, but by a distributed population code, both within primary taste cortex as well as regions involved in processing the hedonic and aversive properties of taste.”
#Pashkovski, S.L., Iurilli, G., Brann, D. et al. Structure and flexibility in cortical representations of odour space. Nature (2020)
https://doi.org/10.1038/s41586-020-2451-1
Quote: “The cortex organizes sensory information to enable discrimination and generalization1,2,3,4. As systematic representations of chemical odour space have not yet been described in the olfactory cortex, it remains unclear how odour relationships are encoded to place chemically distinct but similar odours, such as lemon and orange, into perceptual categories, such as citrus5,6,7. Here, by combining chemoinformatics and multiphoton imaging in the mouse, we show that both the piriform cortex and its sensory inputs from the olfactory bulb represent chemical odour relationships through correlated patterns of activity. However, cortical odour codes differ from those in the bulb: cortex more strongly clusters together representations for related odours, selectively rewrites pairwise odour relationships, and better matches odour perception. The bulb-to-cortex transformation depends on the associative network originating within the piriform cortex, and can be reshaped by passive odour experience. Thus, cortex actively builds a structured representation of chemical odour space that highlights odour relationships; this representation is similar across individuals but remains plastic, suggesting a means through which the olfactory system can assign related odour cues to common and yet personalized percepts.”
– Any moment you perceive is made from different parts. As you are watching this video, your eyes activate columns for vision and color in your visual cortex, your ears transmit information to your auditory cortex, language areas are decoding my words, while other networks keep track of your body and emotions.
#Ichinose T and Habib S (2022) On and off signaling pathways in the retina and the visual system. Front. Ophthalmol.
https://www.frontiersin.org/journals/ophthalmology/articles/10.3389/fopht.2022.989002/full
Quote: “Visual processing starts at the retina of the eye, and signals are then transferred primarily to the visual cortex and the tectum. In the retina, multiple neural networks encode different aspects of visual input, such as color and motion. Subsequently, multiple neural streams in parallel convey unique aspects of visual information to cortical and subcortical regions. Bipolar cells, which are the second-order neurons of the retina, separate visual signals evoked by light and dark contrasts and encode them to ON and OFF pathways, respectively. The interplay between ON and OFF neural signals is the foundation for visual processing for object contrast which underlies higher order stimulus processing. ON and OFF pathways have been classically thought to signal in a mirror-symmetric manner. However, while these two pathways contribute synergistically to visual perception in some instances, they have pronounced asymmetries suggesting independent operation in other cases. In this review, we summarize the role of the ON–OFF dichotomy in visual signaling, aiming to contribute to the understanding of visual recognition.”
#Jia K, Goebel R, Kourtzi Z. Ultra-High Field Imaging of Human Visual Cognition. Annu Rev Vis Sci. 2023
https://www.annualreviews.org/content/journals/10.1146/annurev-vision-111022-123830
Quote: “Functional magnetic resonance imaging (fMRI), the key methodology for mapping the functions of the human brain in a noninvasive manner, is limited by low temporal and spatial resolution. Recent advances in ultra-high field (UHF) fMRI provide a mesoscopic (i.e., submillimeter resolution) tool that allows us to probe laminar and columnar circuits, distinguish bottom-up versus top-down pathways, and map small subcortical areas. We review recent work demonstrating that UHF fMRI provides a robust methodology for imaging the brain across cortical depths and columns that provides insights into the brain's organization and functions at unprecedented spatial resolution, advancing our understanding of the fine-scale computations and interareal communication that support visual cognition.
[...]
Recent advances in ultra-high field (UHF) imaging techniques provide a mesoscopic tool ( Figure 1 ) that enables us to interrogate brain computations at a finer scale (i.e., submillimeter resolution) than that afforded by standard 3T fMRI (Goense et al. 2016). In particular, UHF fMRI allows us to probe cortical layers that are known to have distinct connectivity with upstream and downstream areas, allowing us to discern bottom-up versus top-down pathways and infer the direction of information flow across areas. Furthermore, UHF fMRI enables us to map small subcortical areas [e.g., the lateral geniculate nucleus (LGN), habenula] and columnar organization [e.g., ocular dominance columns (ODCs), orientation columns], allowing us to characterize fine-scale activity that supports intra-areal computations.”
#Ai H, Lin W, Liu C, Chen N, Zhang P. Mesoscale functional organization and connectivity of color, disparity, and naturalistic texture in human second visual area. Elife. 2025
https://pmc.ncbi.nlm.nih.gov/articles/PMC11925451/
Quote: “Although parallel processing has been extensively studied in the low-level geniculostriate pathway and the high-level dorsal and ventral visual streams, less is known at the intermediate-level visual areas. In this study, we employed high-resolution fMRI at 7T to investigate the columnar and laminar organizations for color, disparity, and naturalistic texture in the human secondary visual cortex (V2), and their informational connectivity with lower- and higher-order visual areas. Although fMRI activations in V2 showed reproducible interdigitated color-selective thin and disparity-selective thick ‘stripe’ columns, we found no clear evidence of columnar organization for naturalistic textures. Cortical depth-dependent analyses revealed the strongest color-selectivity in the superficial layers of V2, along with both feedforward and feedback informational connectivity with V1 and V4. Disparity selectivity was similar across different cortical depths of V2, which showed significant feedforward and feedback connectivity with V1 and V3ab. Interestingly, the selectivity for naturalistic texture was strongest in the deep layers of V2, with significant feedback connectivity from V4. Thus, while local circuitry within cortical columns is crucial for processing color and disparity information, feedback signals from V4 are involved in generating the selectivity for naturalistic textures in area V2.”
#Pyott SJ, Pavlinkova G, Yamoah EN, Fritzsch B. Harmony in the Molecular Orchestra of Hearing: Developmental Mechanisms from the Ear to the Brain. Annu Rev Neurosci. 2024
https://www.annualreviews.org/content/journals/10.1146/annurev-neuro-081423-093942
Quote: “Auditory processing in mammals begins in the peripheral inner ear and extends to the auditory cortex. Sound is transduced from mechanical stimuli into electrochemical signals of hair cells, which relay auditory information via the primary auditory neurons to cochlear nuclei. Information is subsequently processed in the superior olivary complex, lateral lemniscus, and inferior colliculus and projects to the auditory cortex via the medial geniculate body in the thalamus. Recent advances have provided valuable insights into the development and functioning of auditory structures, complementing our understanding of the physiological mechanisms underlying auditory processing. This comprehensive review explores the genetic mechanisms required for auditory system development from the peripheral cochlea to the auditory cortex. We highlight transcription factors and other genes with key recurring and interacting roles in guiding auditory system development and organization. Understanding these gene regulatory networks holds promise for developing novel therapeutic strategies for hearing disorders, benefiting millions globally.”
#Moerel M, De Martino F, Uğurbil K, Formisano E, Yacoub E. Evaluating the Columnar Stability of Acoustic Processing in the Human Auditory Cortex. J Neurosci. 2018
https://pmc.ncbi.nlm.nih.gov/articles/PMC6125808/
Quote: “Using ultra-high field fMRI, we explored the cortical depth-dependent stability of acoustic feature preference in human auditory cortex. We collected responses from human auditory cortex (subjects from either sex) to a large number of natural sounds at submillimeter spatial resolution, and observed that these responses were well explained by a model that assumes neuronal population tuning to frequency-specific spectrotemporal modulations. We observed a relatively stable (columnar) tuning to frequency and temporal modulations. However, spectral modulation tuning was variable throughout the cortical depth. This difference in columnar stability between feature maps could not be explained by a difference in map smoothness, as the preference along the cortical sheet varied in a similar manner for the different feature maps. Furthermore, tuning to all three features was more columnar in primary than nonprimary auditory cortex. The observed overall lack of overlapping columnar regions across acoustic feature maps suggests, especially for primary auditory cortex, a coding strategy in which across cortical depths tuning to some features is kept stable, whereas tuning to other features systematically varies.”
#F. De Martino, M. Moerel, K. Ugurbil, R. Goebel, E. Yacoub, & E. Formisano, Frequency preference and attention effects across cortical depths in the human primary auditory cortex, Proc. Natl. Acad. Sci. U.S.A. (2015)
https://doi.org/10.1073/pnas.1507552112
Quote: “Columnar arrangements of neurons with similar preference have been suggested as the fundamental processing units of the cerebral cortex. Within these columnar arrangements, feed-forward information enters at middle cortical layers whereas feedback information arrives at superficial and deep layers. This interplay of feed-forward and feedback processing is at the core of perception and behavior. Here we provide in vivo evidence consistent with a columnar organization of the processing of sound frequency in the human auditory cortex. We measure submillimeter functional responses to sound frequency sweeps at high magnetic fields (7 tesla) and show that frequency preference is stable through cortical depth in primary auditory cortex. Furthermore, we demonstrate that—in this highly columnar cortex—task demands sharpen the frequency tuning in superficial cortical layers more than in middle or deep layers. These findings are pivotal to understanding mechanisms of neural information processing and flow during the active perception of sounds.”
#Hamilton LS, Oganian Y, Hall J, Chang EF. Parallel and distributed encoding of speech across human auditory cortex. Cell. 2021
https://pmc.ncbi.nlm.nih.gov/articles/PMC8456481/
Quote: “Speech perception is thought to rely on a cortical feedforward serial transformation of acoustic into linguistic representations. Using intracranial recordings across the entire human auditory cortex, electrocortical stimulation, and surgical ablation, we show that cortical processing across areas is not consistent with a serial hierarchical organization. Instead, response latency and receptive field analyses demonstrate parallel and distinct information processing in the primary and nonprimary auditory cortices. This functional dissociation was also observed where stimulation of the primary auditory cortex evokes auditory hallucination but does not distort or interfere with speech perception. Opposite effects were observed during stimulation of nonprimary cortex in superior temporal gyrus. Ablation of the primary auditory cortex does not affect speech perception. These results establish a distributed functional organization of parallel information processing throughout the human auditory cortex and demonstrate an essential independent role for nonprimary auditory cortex in speech processing.”
#Forseth, K.J., Hickok, G., Rollo, P.S. et al. Language prediction mechanisms in human auditory cortex. Nat Commun (2020)
https://doi.org/10.1038/s41467-020-19010-6
Quote: “Spoken language, both perception and production, is thought to be facilitated by an ensemble of predictive mechanisms. We obtain intracranial recordings in 37 patients using depth probes implanted along the anteroposterior extent of the supratemporal plane during rhythm listening, speech perception, and speech production. These reveal two predictive mechanisms in early auditory cortex with distinct anatomical and functional characteristics. The first, localized to bilateral Heschl’s gyri and indexed by low-frequency phase, predicts the timing of acoustic events. The second, localized to planum temporale only in language-dominant cortex and indexed by high-gamma power, shows a transient response to acoustic stimuli that is uniquely suppressed during speech production. Chronometric stimulation of Heschl’s gyrus selectively disrupts speech perception, while stimulation of planum temporale selectively disrupts speech production. This work illuminates the fundamental acoustic infrastructure—both architecture and function—for spoken language, grounding cognitive models of speech perception and production in human neurobiology.”
#Emma J.P. Brouwer, Nikos Priovoulos, Julie Hashimoto, Wietske van der Zwaag; Proprioceptive engagement of the human cerebellum studied with 7T-fMRI. Imaging Neuroscience. 2024
Quote: “Our ability to sense our bodies in space—referred to as proprioception (Bhanpuri et al., 2013)—is indispensable for every-day tasks such as maintaining posture, balancing on one leg, or simply reaching for a cup of coffee (Bloem et al., 2002). While we perform these tasks without much thought, the involvement of many brain regions is required. Proprioceptive input travels through the dorsal root ganglia through the posterior columns of the spinal cord and cuneate nucleus, through the thalamus to the primary and secondary motorsensory cortices and the cerebellum, informing these regions on the position and forces of muscles (Goble et al., 2012; Goossens et al., 2019; Iandolo et al., 2018; Marin Vargas et al., 2024; Smoth & Deacon, 1984). Although the cerebellum is known to be important for proprioceptive processing, its functional organisation is much less explored than that of the cerebral cortex. This knowledge gap arises at least partly from the inherent challenges posed by the cerebellar cortex’s intricate and highly folded structure, necessitating sophisticated high-resolution acquisition and analysis methods.”
#Palomero-Gallagher, N., Amunts, K. A short review on emotion processing: a lateralized network of neuronal networks. Brain Struct Funct (2022) https://doi.org/10.1007/s00429-021-02331-7
Quote: “Emotions are valenced mental responses and associated physiological reactions that occur spontaneously and automatically in response to internal or external stimuli, and can influence our behavior, and can themselves be modulated to a certain degree voluntarily or by external stimuli. They are subserved by large-scale integrated neuronal networks with epicenters in the amygdala and the hippocampus, and which overlap in the anterior cingulate cortex. Although emotion processing is accepted as being lateralized, the specific role of each hemisphere remains an issue of controversy, and two major hypotheses have been proposed. In the right-hemispheric dominance hypothesis, all emotions are thought to be processed in the right hemisphere, independent of their valence or of the emotional feeling being processed. In the valence lateralization hypothesis, the left is thought to be dominant for the processing of positively valenced stimuli, or of stimuli inducing approach behaviors, whereas negatively valenced stimuli, or stimuli inducing withdrawal behaviors, would be processed in the right hemisphere. More recent research points at the existence of multiple interrelated networks, each associated with the processing of a specific component of emotion generation, i.e., its generation, perception, and regulation. It has thus been proposed to move from hypotheses supporting an overall hemispheric specialization for emotion processing toward dynamic models incorporating multiple interrelated networks which do not necessarily share the same lateralization patterns.”
– All these signals are processed in deeper areas of your brain that evaluate them, boosting what seems important right now and tuning out what doesn’t.
This is an extremely complex process, but in a nutshell: the brain filters the constant flood of sensory inputs using selective attention, boosting relevant signals and suppressing irrelevant ones. For example, in visual perception and processing, sensory and brain structures compute so-called “salience maps” based on basic features (brightness, contrast, etc) and context. These maps then support “competitive selection”: stronger or more relevant stimuli suppress weaker ones, so only the “highest priority” features dominate processing.
#Knudsen EI. Neural Circuits That Mediate Selective Attention: A Comparative Perspective. Trends Neurosci. 2018
https://pmc.ncbi.nlm.nih.gov/articles/PMC6204111/
Quote: “Since the beginning of vertebrate evolution, neural mechanisms of attention have selected the information that gains access to networks that make cognitive decisions. When an animal, be it fish or primate, is engaged in complex behavior, such as social interactions, navigation or foraging, information selection is based on the task and goals of the animal [1–3]. During such periods, the mechanisms of attention are controlled by forebrain networks [3–7]. However, when an unexpected or highly salient stimulus occurs, information selection is dominated by the physical properties of the stimulus. When the stimulus has a location, a midbrain network acts with speed to direct spatial attention to that location and, when appropriate, also the gaze of the animal [1, 8–10]. Following capture of spatial attention by a physically salient stimulus, forebrain networks identify and evaluate the risks and benefits of the stimulus.
The forebrain and midbrain networks that mediate these complementary aspects of attention-control each contain specialized circuits that compute the highest priority information at each moment for decision-making (Fig. 1, Key Figure). Forebrain networks select information, based on task demands or the physical salience of stimuli, from all available sources, including sensory input, plans for action, and memory stores. They direct attention either to locations, sensory modalities, stimulus features, objects or memory stores [4]. In contrast, the midbrain network is concerned only with the relative priorities of locations (Fig. 1A, purple box), based on the physical salience of stimuli and their behavioral relevance, and directs spatial attention to the highest priority location [1, 8, 11, 12].”
– Vastly simplifying, this concert of all these different gears connected by levers, wheels and screws comes together to create a new structure, within your brain machine: An assembly. A new pattern of synchronized activity within the grand architecture of your brain.
Cell assemblies are groups of neurons that tend to fire together within very tight time windows (millisecond scale). This near-synchronous firing is (among other things) central to how we experience the world and how memories are encoded in the brain and recalled.
Experiencing the current moment is thought to rely on stimulus‑specific neuronal assemblies that transiently activate in the sensory cortex and related circuits in the brain. These assemblies are essentially considered to be the basic units of perception, where each assembly encodes a concept or feature.
#Miehl, C., Onasch, S., Festa, D. and Gjorgjieva, J. Formation and computational implications of assemblies in neural circuits. J Physiol. (2023)
https://doi.org/10.1113/JP282750
Quote: “In the brain, patterns of neural activity represent sensory information and store it in non-random synaptic connectivity. A prominent theoretical hypothesis states that assemblies, groups of neurons that are strongly connected to each other, are the key computational units underlying perception and memory formation. Compatible with these hypothesised assemblies, experiments have revealed groups of neurons that display synchronous activity, either spontaneously or upon stimulus presentation, and exhibit behavioural relevance. While it remains unclear how assemblies form in the brain, theoretical work has vastly contributed to the understanding of various interacting mechanisms in this process. Here, we review the recent theoretical literature on assembly formation by categorising the involved mechanisms into four components: synaptic plasticity, symmetry breaking, competition and stability. We highlight different approaches and assumptions behind assembly formation and discuss recent ideas of assemblies as the key computational unit in the brain.”
#Holtmaat, A., Caroni, P. Functional and structural underpinnings of neuronal assembly formation in learning. Nat Neurosci (2016)
https://doi.org/10.1038/nn.4418
Quote: “Learning and memory are associated with the formation and modification of neuronal assemblies: populations of neurons that encode what has been learned and mediate memory retrieval upon recall. Functional studies of neuronal assemblies have progressed dramatically thanks to recent technological advances. Here we discuss how a focus on assembly formation and consolidation has provided a powerful conceptual framework to relate mechanistic studies of synaptic and circuit plasticity to behaviorally relevant aspects of learning and memory. Neurons are likely recruited to particular learning-related assemblies as a function of their relative excitabilities and synaptic activation, followed by selective strengthening of pre-existing synapses, formation of new connections and elimination of outcompeted synapses to ensure memory formation. Mechanistically, these processes involve linking transcription to circuit modification. They include the expression of immediate early genes and specific molecular and cellular events, supported by network-wide activities that are shaped and modulated by local inhibitory microcircuits.”
#Filipchuk, A., Schwenkgrub, J., Destexhe, A. et al. Awake perception is associated with dedicated neuronal assemblies in the cerebral cortex. Nat Neurosci (2022)
https://doi.org/10.1038/s41593-022-01168-5
Quote: “Neural activity in the sensory cortex combines stimulus responses and ongoing activity, but it remains unclear whether these reflect the same underlying dynamics or separate processes. In the present study, we show in mice that, during wakefulness, the neuronal assemblies evoked by sounds in the auditory cortex and thalamus are specific to the stimulus and distinct from the assemblies observed in ongoing activity. By contrast, under three different anesthetics, evoked assemblies are indistinguishable from ongoing assemblies in the cortex. However, they remain distinct in the thalamus. A strong remapping of sensory responses accompanies this dynamic state change produced by anesthesia. Together, these results show that the awake cortex engages dedicated neuronal assemblies in response to sensory inputs, which we suggest is a network correlate of sensory perception.”
#Langille JJ and Brown RE. The Synaptic Theory of Memory: A Historical Survey and Reconciliation of Recent Opposition. Front. Syst. Neurosci. (2018)
https://www.frontiersin.org/journals/systems-neuroscience/articles/10.3389/fnsys.2018.00052/full
Quote: “Hebb’s (1949) theory postulated that the neurophysiological changes underlying learning and memory occur in three stages: (1) synaptic changes; (2) formation of a “cell assembly”; and (3) formation of a “phase sequence,” which link the neurophysiological changes underlying learning and memory as studied by physiologists to the study of thought and “mind” as conceived by cognitive psychologists. As pointed out by Trettenbrein (2016), Hebb’s neurophysiological postulate (Hebb, 1949, page 62) states that:
“When an axon of cell A is near enough to excite a cell B and repeatedly or persistently takes part in firing it, some growth process or metabolic change takes place in one or both cells such that A’s efficiency, as one of the cells firing B, is increased.”
The “cell assembly” (Hebb, 1949, pages 69–74) is a set of neurons and the pathways connecting them, which act together, such that a stimulus activating pathway 1 will activate a reverberating circuit of N pathways (in Hebb’s example, n = 15). A cell assembly is a hypothetical reverberating system, proposed as a mediating process, an element of thought, capable of holding an excitation and thus of bridging a gap in time between stimulus and response (Hebb, 1972, pages 295 and 304). A series of cell assemblies, connected by neural activity over time is a “Phase Sequence,” which provides the neural basis for a “train of thought” from one cell assembly to another (Hebb, 1949, pages 79–106). The cell assembly thus “relates the individual nerve cell to psychological phenomenon” such that “a bridge has been thrown across the great gap between the details of neurophysiology and the molar conceptions of psychology” (Hebb, 1949, page 101). Hebb then elaborated on how this theory could account for learning and memory, how new learning could be associated with previous learning, and how “quick learning” (perhaps similar to the single trial learning of Gallistel and Balsam (2014)) might occur (Hebb, 1949, chapter 8). Hebb’s cell assembly theory thus showed how differences between psychologists and physiologists, who often use different definitions for the same phenomena, could be reconciled into a theory of the neurophysiological basis of learning and memory. It is important to note that Hebb’s postulate quoted above contains two concepts: synaptic plasticity and “some growth process or metabolic change” in the neuron, which has been termed “intrinsic plasticity” (Sehgal et al., 2013; Titley et al., 2017).”
– In reality there is not just one assembly active in your brain. Your brain can’t process everything in the outer and inner world with full attention, so different assemblies are fighting for dominance. The details are complicated, but at any moment, one assembly is winning and deemed the most important by your brain. This is what you are aware of right now.
Neural representations of external stimuli are filtered and processed so that we consciously perceive only some of them. Prominent theories on neurocognition posit that these neural representations are interpretable as cell assemblies (groups of neurons that are highly connected). The details of how this works are not fully understood and we are simplifying a lot here. But in a nutshell, these neural representations are thought to “compete” for attention in the brain by e.g. suppressing each other and amplifying themselves. The result is a winner-take-most outcome, where one cell assembly (a group of highly connected neurons) or other neural representation dominates processing and behavior.
#Aguado-López B, Palenciano AF, Peñalver JMG, Díaz-Gutiérrez P, López-García D, Avancini C, Ciria LF, Ruz M. Proactive selective attention across competition contexts. Cortex. 2024
https://www.sciencedirect.com/science/article/pii/S001094522400114X
Quote: “In our everyday life we are surrounded by myriads of stimuli, but only some of them occupy our mind. The process of selective attention relates to the filtering of unwanted information and the selection of the pieces that are relevant to us. This intricate cognitive function proceeds through various biasing routes: one involves bottom-up processes, where the characteristics of stimuli automatically capture attention, while another is guided by goals, following top-down processes (Desimone & Duncan, 1995). The latter does not only take place during stimulus processing but can also happen in anticipatory fashion, by activating goal-related information before target events. This preparatory selection is related to proactive cognition (Braver, 2012). However, research about how such preparation unfolds to aid selection in contexts with different attentional demands is scarce.
One of the most influential proposals explaining selective attention is the biased competition model (Desimone & Duncan, 1995). This framework highlights the limited capacity of neural information processing and the essential role of competition in resolving this problem. As we move up along the cortical hierarchy, neurons increase their receptive fields to respond to stimuli. However, given the limits of their response, the more pieces of information (e.g., objects) are placed in the same receptive field, the less information there will be about each of them. Thus, neurons are selective and prioritize certain types of information, and thus the information that reaches our senses competes to be represented. Such competition is biased by bottom-up and top-down mechanisms. While bottom-up mechanisms may favor, for example, the most salient stimulus, top-down processes engage neurons in the prefrontal cortex that generate internal templates representing relevant information. These templates are used to bias neural competition favoring goal-relevant information. For example, if we aim to recognize a friend's face in a crowd, pre-activation of a template of that face would later guide the attentional selection.”
#Shreesh P Mysore, Ninad B Kothari. Mechanisms of competitive selection: A canonical neural circuit framework. eLife (2020)
https://elifesciences.org/articles/51473
Quote: “Competitive selection, the transformation of multiple competing sensory inputs and internal states into a unitary choice, is a fundamental component of animal behavior. Selection behaviors have been studied under several intersecting umbrellas including decision-making, action selection, perceptual categorization, and attentional selection. Neural correlates of these behaviors and computational models have been investigated extensively. However, specific, identifiable neural circuit mechanisms underlying the implementation of selection remain elusive. Here, we employ a first principles approach to map competitive selection explicitly onto neural circuit elements. We decompose selection into six computational primitives, identify demands that their execution places on neural circuit design, and propose a canonical neural circuit framework. The resulting framework has several links to neural literature, indicating its biological feasibility, and has several common elements with prominent computational models, suggesting its generality. We propose that this framework can help catalyze experimental discovery of the neural circuit underpinnings of competitive selection.”
#Stöber TM, Lehr AB, Nikzad A, Ganjtabesh M, Fyhn M, et al. (2025) Competition and cooperation of assembly sequences in recurrent neural networks. PLOS Computational Biology
https://doi.org/10.1371/journal.pcbi.1013403
Quote: “Neural activity sequences are ubiquitous in the brain and play pivotal roles in functions such as long-term memory formation and motor control. While conditions for storing and reactivating individual sequences have been thoroughly characterized, it remains unclear how multiple sequences may interact when activated simultaneously in recurrent neural networks. This question is especially relevant for weak sequences, comprised of fewer neurons, competing against strong sequences. Using a non-linear rate -based and a spiking model with discrete, pre-configured assemblies, we demonstrate that weak sequences can compensate for their competitive disadvantage either by increasing excitatory connections between subsequent assemblies or by cooperating with other co-active sequences. Further, our models suggest that such cooperation can negatively affect sequence speed unless subsequently active assemblies are paired. Our analysis characterizes the conditions for successful sequence progression in isolated, competing, and cooperating assembly sequences, and identifies the distinct contributions of recurrent and feed-forward projections. This proof-of-principle study shows how even disadvantaged sequences can be prioritized for reactivation, a process which has recently been implicated in hippocampal memory processing.”
– Two things are happening right now: The neurons of the winning assembly are bathed in chemicals that make them more susceptible to change and tie them closer together, strengthening the synapses between them.
Here we are referencing several related concepts relating to the brain's ability to change and adapt by modifying the strength of connections (synapses) between neurons (synaptic plasticity). All of these processes are crucial for learning and memory. “Hebbian learning” is when a “winning” cell assembly (a group of highly connected neurons) is repeatedly active, synapses between its neurons are strengthened and sometimes multiplied.
Your brain also makes molecules called “neuromodulators (e.g. dopamine, adrenaline, serotonin. They have many effects on the brain, and can also make synapses more likely to get strengthened in response to recent neuronal activity. This strengthening, called long-term potentiation, can involve both increased synaptic efficacy and, over longer timescales, the formation of additional synapses.
#Magee JC, Grienberger C. Synaptic Plasticity Forms and Functions. Annu Rev Neurosci. 2020
https://www.annualreviews.org/content/journals/10.1146/annurev-neuro-090919-022842
Quote: “One of the most influential ideas about how learning-related changes might occur in brains was first articulated by Donald Hebb (1949) around seventy years ago. Hebb's famous postulate states that if cell A “repeatedly or persistently takes part” in firing cell B, then the strength of their connection should increase. In this statement, Hebb points out that learning-related changes should be found in the strength of the synaptic connections between individual neurons within a neuronal population or assembly and that these changes should, in his view, be based on causality and repetition. The underlying ideas here are that synapses responsible for correct responses (reward, escape, etc.) should be properly credited (this is known as credit assignment) and that this process should be gradual to avoid the errors associated with noise.
In the following decades, the pioneers of artificial neural networks (ANNs) interpreted Hebb's idea to be that weight changes among the units of single-layer networks should be based on coincidence, or the product of pre- and postsynaptic activity, thus transforming the causality into changes proportional to the coactivity or correlation of the input and output units (Rosenblatt 1959, Woodrow & Hoff 1960) ( Figure 1a ). The most straightforward of these network learning rules, the unsupervised forms, produces a rudimentary form of memory by linking a system's particular input patterns with the output patterns that are consistently associated with them. Here, inputs driving a given output will have their connection strengths enhanced, forming a simple association that allows even fragments of the associated input pattern to evoke the correct output activity (Andersen 1972, Kohonen 1972, Hopfield 1982). In addition, these unsupervised Hebbian learning rules autonomously find low-order statistical structure in the input stream, allowing them to mediate some types of fundamental feature selectivity and to generate self-organizing topographical representations of those features (von der Malsberg 1973, Grossberg 1976, Kohonen 1982, Oja 1982, Erwin & Miller 1998, Song & Abbott 2001, Brito & Gerstner 2016).”
#Foncelle A, Mendes A, Jędrzejewska-Szmek J, Valtcheva S, Berry H, Blackwell KT and Venance L. Modulation of Spike-Timing Dependent Plasticity: Towards the Inclusion of a Third Factor in Computational Models. Front. Comput. Neurosci. (2018)
Quote: “Most computational and experimental studies of synaptic plasticity focus on variations of Hebb’s rule in which the change in synaptic strength is caused by direct association of two factors, i.e., two inputs (or activity patterns), one on the presynaptic and one on the postsynaptic side. Thus, when neural circuits adjust their synaptic weights depending on the frequency or timing of the pre-synaptic and post-synaptic firing patterns, Hebb’s postulate is fulfilled. In addition, a third factor (for example neuromodulators or astrocytes) stabilizes or modulates the expression of synaptic plasticity and, thus, ultimately learning (Kempter et al., 1998; Pawlak et al., 2010; Lisman et al., 2011; Frémaux and Gerstner, 2016; Edelmann et al., 2017; Kuśmierz et al., 2017; Gerstner et al., 2018). The inclusion of this third factor with two-factor Hebbian plasticity rule is called neoHebbian plasticity (Lisman et al., 2011), and is infrequent in computational models of spike-timing dependent plasticity (STDP). In this review article, we focus on STDP (Sjöström et al., 2008; Feldman, 2012), a synaptic Hebbian learning rule, and its control by the third factor: neuromodulation (via the action of dopamine, acetylcholine, noradrenaline and others) or astrocyte activity. Our goal is to highlight aspects of STDP that should be taken into account in future computational models of STDP.”
#Debanne D, Inglebert Y. Spike timing-dependent plasticity and memory. Curr Opin Neurobiol. 2023
https://pubmed.ncbi.nlm.nih.gov/36924615/
Quote: “Spike timing-dependent plasticity (STDP) is a bidirectional form of synaptic plasticity discovered about 30 years ago and based on the relative timing of pre- and post-synaptic spiking activity with a millisecond precision. STDP is thought to be involved in the formation of memory but the millisecond-precision spike-timing required for STDP is difficult to reconcile with the much slower timescales of behavioral learning. This review therefore aims to expose and discuss recent findings about i) the multiple STDP learning rules at both excitatory and inhibitory synapses in vitro, ii) the contribution of STDP-like synaptic plasticity in the formation of memory in vivo and iii) the implementation of STDP rules in artificial neural networks and memristive devices.”
#Ang GWY, Tang CS, Hay YA, Zannone S, Paulsen O, et al. The functional role of sequentially neuromodulated synaptic plasticity in behavioural learning. PLOS Computational Biology (2021)
https://doi.org/10.1371/journal.pcbi.1009017
Quote: “When the environment changes and previous reward associations no longer hold, an animal must quickly adapt its behaviour to maximize reward. The learning rules in the brain responsible for updating action-outcome contingencies in such situations are not fully understood. Traditional forms of Hebbian plasticity [1, 2], including spike-timing-dependent-plasticity (STDP) [3–7], change synaptic weights based on the joint activation of pre- and post- synaptic neurons alone. They do not account for behavioural learning paradigms that require external feedback. Synaptic plasticity that is regulated by neuromodulators [8–11] provides a mechanism to incorporate behaviourally relevant information into synaptic changes, and at the appropriate time. Neuromodulatory signals are released in response to certain salient events (e.g. reward discovery or reward removal) and gate plasticity, depressing or potentiating recently active synapses responsible for the outcome, changing behaviour in a task relevant way [12].”
#Lehr, A.B., Luboeinski, J. & Tetzlaff, C. Neuromodulator-dependent synaptic tagging and capture retroactively controls neural coding in spiking neural networks. Sci Rep (2022).
https://doi.org/10.1038/s41598-022-22430-7
Quote: “It is commonly thought that neural activity changes synaptic connections to encode the memory of an experience1,2,3. Recall of this memory is, in turn, considered to depend on the same subset of neurons becoming active again4,5,6,7. Thus, characterizing neural activity during the initial experience and during recall should reveal the neural code that supports memory. As yet, the defining features of neural activity required for neural computation, in particular memory recall, remain a topic of active investigation. On the one hand, information in the brain may be subject to so-called rate-based coding, meaning that information is conveyed by the number of spikes in a particular time interval. On the other hand, it may be subject to temporal (sometimes called spike-based) coding, meaning that the information is conveyed by the timing of individual spikes8,9. Evidence for either scheme has been found in many experimental studies and the two schemes can be expressed to a different degree depending on task, brain region, and experimental variables10,11,12,13,14. For instance in macaque somatosensory cortex, the amplitude of tactile vibrations is encoded in firing rates, while the frequency of vibrations is encoded in precise spike patterns15. In rat primary and secondary somatosensory cortices, information about textured surfaces is conveyed by both rate and spike time codes carrying approximately independent, complementary information about the stimulus16. In crustacean mechanoreceptors, the occurrence of precise spike-timing and robust rate coding has been shown to be negatively correlated in a neuromodulator-dependent manner, with serotonin favoring higher rates and allatostatin favoring precise spike-timing17. Across a number of species, hippocampal and entorhinal cortex neurons, crucial for semantic and episodic memory18, exhibit firing patterns indicative of both rate and spike-based coding schemes19,20,21. These coding schemes have been shown to vary independently11 and the rate code can remain preserved even if the spike-based code is disrupted22,23. Furthermore, hippocampal coding can be task-dependent, as temporally precise activity was observed during a working memory task with a temporal component but was absent in non-memory control tasks24. Taken together, in both sensory and memory systems, the extent of rate-based and spike-based coding flexibly depends on brain state and task. [...] It is known that neuromodulators must be present for successful memory formation28,29,30 as well as for initial memory consolidation (synaptic consolidation), for which there is a time window of up to hours in which neuromodulation has an effect31,32,33,34,35,36. Neuromodulator signals can even retroactively affect a memory trace in the minutes to hours after the related event took place27,37,38,39,40. Thus, the effect of selective neuromodulation on consolidation makes it an ideal candidate for retroactively changing the neural representation of an experience.”
#Borczyk M, Radwanska K, Giese KP. The importance of ultrastructural analysis of memory. Brain Res Bull. 2021
https://www.sciencedirect.com/science/article/pii/S0361923021001222?via%3Dihub
Quote: “Plasticity of glutamatergic synapses in the hippocampus is believed to underlie learning and memory processes. Surprisingly, very few studies report long-lasting structural changes of synapses induced by behavioral training. It remains, therefore, unclear which synaptic changes in the hippocampus contribute to memory storage. Here, we systematically compare how long-term potentiation of synaptic transmission (LTP) (a primary form of synaptic plasticity and cellular model of memory) and behavioral training affect hippocampal glutamatergic synapses at the ultrastructural level enabled by electron microscopy. The review of the literature indicates that while LTP induces growth of dendritic spines and post-synaptic densities (PSD), that represent postsynaptic part of a glutamatergic synapse, after behavioral training there is transient (< 6 h) synaptogenesis and long-lasting (> 24 h) increase in PSD volume (without a significant change of dendritic spine volume), indicating that training-induced PSD growth may reflect long-term enhancement of synaptic functions. Additionally, formation of multi-innervated spines (MIS), is associated with long-term memory in aged mice and LTP-deficient mutant mice. Since volume of PSD, as well as atypical synapses, can be reliably observed only with electron microscopy, we argue that the ultrastructural level of analysis is required to reveal synaptic changes that are associated with long-term storage of information in the brain.”
#Palacios-Filardo J, Mellor JR. Neuromodulation of hippocampal long-term synaptic plasticity. Curr Opin Neurobiol. 2019
https://pmc.ncbi.nlm.nih.gov/articles/PMC6367596/
Quote: “
Acetylcholine, noradrenaline, dopamine and serotonin all facilitate long-term synaptic plasticity.
Neuromodulators facilitate long-term synaptic plasticity by common and divergent mechanisms.
Common mechanisms include NMDA receptor facilitation by potassium channel inhibition, gliotransmission and disinhibition.
Divergent mechanisms include diversity of disinhibition and temporal and spatial neuromodulator release.”
– And the hippocampus, the memory center and librarian of your brain is activated. The details here are super complicated and the exact process isn't fully understood yet. But in a nutshell, your hippocampus creates a blueprint, saving the rough configuration of the assembly. The assembly is saved and put on an index with all your other memories that are associated with what was going on in this moment.
The hippocampus is central for forming and organizing episodic memories (memories of events in time and space) and spatial memory, for e.g. navigation. It is not just a simple “memory store”, but instead it encodes, binds, replays, and reorganizes event information in space and time, enabling episodic memory, navigation, and planning.
While the importance of the hippocampus for memory is widely established, the exact mechanism of how memories are formed and organized is an active area of research. Hippocampal “cell assemblies” (groups of neurons that are highly connected) are thought to act as indices that link together widely distributed neural activity patterns making up an episode.
More specifically, Memory Indexing Theory proposes that engrams, i.e. the physical or biochemical changes in the brain representing a memory, in the hippocampus do not store all memory content but form a sparse index code that binds together the neural patterns active during an experience.
Cell assemblies associated with memories are also stored in the hippocampus. This means that the hippocampus does not just “index” the memories, it also stores some itself.
We are also simplifying a lot here. It takes hours to store a memory for several days, and even days-to-weeks to store it for longer or permanently. The first learning event triggers a cascade of cellular and molecular processes that must happen for a memory to be stored. If this does not happen, it only lasts as working memory or short-term memory (for minutes-hours). In this video we only give a general overview of the broad strokes process.
#Kolibius LD, Josselyn SA, Hanslmayr S. On the origin of memory neurons in the human hippocampus. Trends Cogn Sci. 2025
https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(25)00031-2
Quote: “The hippocampus is essential for episodic memory, yet its coding mechanism remains debated. In humans, two main theories have been proposed: one suggests that concept neurons represent specific elements of an episode, while another posits a conjunctive code, where index neurons code the entire episode. Here, we integrate new findings of index neurons in humans and other animals with the concept-specific memory framework, proposing that concept neurons evolve from index neurons through overlapping memories. This process is supported by engram literature, which posits that neurons are allocated to a memory trace based on excitability and that reactivation induces excitability. By integrating these insights, we connect two historically disparate fields of neuroscience: engram research and human single neuron episodic memory research.
[...]
How the hippocampus codes episodic memories
Remember the last time you had coffee with a friend in your favorite coffee place. Loungy jazz music played in the background as a waiter attended another table. Episodic memory (see Glossary) [1,2] enables us to recall such past multimodal experiences with remarkable detail even after a single exposure. This process is fundamentally tied to the hippocampus. However, recent findings in humans are contradictory regarding how neurons in the hippocampus implement this episodic code. Current literature offers two possible explanations: one framework based on concept neurons coding specific elements within an episode [3–5] and an alternative based on index neurons coding the conjunction of these elements [6,7]. In this opinion, we first examine concept neurons in more depth, emphasizing their role as potential building blocks of episodic memory. We then explore index neurons, highlighting recent findings in humans and other species that provide strong support for a conjunctive coding framework. Next, we connect these two traditionally separate research streams by integrating insights from engram theory, specifically focusing on neuronal allocation via excitability. Building on this foundation, we propose a unified model in which index neurons gradually transform into concept neurons through overlapping memory traces.”
#Barnett AJ, Reilly W, Dimsdale-Zucker HR, Mizrak E, Reagh Z, Ranganath C. Intrinsic connectivity reveals functionally distinct cortico-hippocampal networks in the human brain. PLoS Biol. 2021
https://pmc.ncbi.nlm.nih.gov/articles/PMC8202937/
Quote: “Episodic memory depends on interactions between the hippocampus and interconnected neocortical regions. Here, using data-driven analyses of resting-state functional magnetic resonance imaging (fMRI) data, we identified the networks that interact with the hippocampus—the default mode network (DMN) and a “medial temporal network” (MTN) that included regions in the medial temporal lobe (MTL) and precuneus. We observed that the MTN plays a critical role in connecting the visual network to the DMN and hippocampus. The DMN could be further divided into 3 subnetworks: a “posterior medial” (PM) subnetwork comprised of posterior cingulate and lateral parietal cortices; an “anterior temporal” (AT) subnetwork comprised of regions in the temporopolar and dorsomedial prefrontal cortex; and a “medial prefrontal” (MP) subnetwork comprised of regions primarily in the medial prefrontal cortex (mPFC). These networks vary in their functional connectivity (FC) along the hippocampal long axis and represent different kinds of information during memory-guided decision-making. Finally, a Neurosynth meta-analysis of fMRI studies suggests new hypotheses regarding the functions of the MTN and DMN subnetworks, providing a framework to guide future research on the neural architecture of episodic memory.”
#Goode TD, Tanaka KZ, Sahay A, McHugh TJ. An Integrated Index: Engrams, Place Cells, and Hippocampal Memory. Neuron. 2020
https://pmc.ncbi.nlm.nih.gov/articles/PMC7486247/
Quote: “The hippocampus and its extended network contribute to encoding and recall of episodic experiences. Drawing from recent anatomical, physiological, and behavioral studies, we propose that hippocampal engrams function as indices to mediate memory recall. We broaden this idea to discuss potential relationships between engrams and hippocampal place cells, as well as the molecular, cellular, physiological, and circuit determinants of engrams that permit flexible routing of information to intra- and extrahippocampal circuits for reinstatement, a feature critical to memory indexing. Incorporating indexing into frameworks of memory function opens new avenues of study and even therapies for hippocampal dysfunction.”
#Tanaka KZ, McHugh TJ. The Hippocampal Engram as a Memory Index. Journal of Experimental Neuroscience. 2018
https://journals.sagepub.com/doi/10.1177/1179069518815942
Quote: “The hippocampus encodes memories for past events, but the nature of the hippocampal code subserving this function remains unclear. A prevailing idea, strongly supported by hippocampal physiology, is the Cognitive Map Theory. In this view, episodic memories are anchored to spatial domains, or allocentric frameworks, of experiences, with the hippocampus providing a stable representation of external space. On the other hand, recent studies using Immediate Early Genes (IEGs) as a proxy of neuronal activation support the Memory Index Theory. This idea posits that the hippocampal memory trace serves as an index for a cortical representation of memory (a map for internal representation) and hypothesizes the primary hippocampal function is to reinstate the pattern of cortical activity present during encoding. Our recent findings provide a unitary view on these two fundamentally different theories. In the hippocampal CA1 region the activity of c-Fos expressing pyramidal neurons reliably reflects the identity of the context the animal is experiencing in an index-like fashion, while spikes from other active pyramidal cells provide spatial information that is stable over a long period of time. These two distinct ensembles of hippocampal neurons suggest heterogeneous roles for subsets of hippocampus neurons in memory.”
#Lorena Deuker, Jacob LS Bellmund, Tobias Navarro Schröder, Christian F Doeller. An event map of memory space in the hippocampus. eLife (2016)
https://elifesciences.org/articles/16534
Quote: “The hippocampus is one of the most extensively studied regions in the brain. However, two of its core functions, spatial navigation and episodic memory, have mostly been investigated in separate research lines (Eichenbaum, 2014). It has been suggested that the answer to the apparent duality in hippocampal function resides in a common mechanism that is required for both spatial navigation and episodic memory (Eichenbaum, 2014): the formation of an abstract representation of the external world, a memory space (Eichenbaum et al., 1999). While it is clear that such a map-like representation would be necessary for spatial navigation, it might be less obvious for episodic memory. Yet, episodic memory has been defined as the ability to recall events from one’s own life (Tulving, 1983) in a specific mode of retrieval that has been referred to as recollection (Eichenbaum et al., 2007) or 'mental time travel' (Tulving, 2002). This specific mode of retrieval makes it necessary that humans can, in their minds, re-create and re-experience episodes of their past by mentally navigating to the point when and where the episode happened, thereby retrieving the time and the place of past events. Notably, this implies that humans must be able to convert relationships between events, for example along the physical dimensions of space and time, into a mental representation so that the arrangement of events is appropriately reflected. In line with this idea, recent discoveries in rodent electrophysiology indicate that cells in the hippocampus code for events in space and time simultaneously (Kraus et al., 2013, 2015; Mankin et al., 2012) and provide evidence for the notion that memories are, in fact, stored in a multi-dimensional memory space (Eichenbaum et al., 1999; McKenzie et al., 2014). Findings from fMRI studies in humans also suggest that memories are dynamically integrated into mnemonic network representations along different dimensions (Collin et al., 2015; Horner et al., 2015; Kumaran and Maguire, 2006; Milivojevic et al., 2015; Preston and Eichenbaum, 2013; Schlichting et al., 2015; Shohamy and Wagner, 2008; Zeithamova et al., 2012). However, it remains elusive how inter-event relationships along multiple dimensions, such as space and time, are combined and converted into a multi-dimensional mnemonic event map, which might potentially support episodic memory.”
– An activation pattern of millions of neurons, spanning many different regions of your brain.
The neural representations of memory are commonly referred to as “engram”. Engram cells are then those assembly neurons that are necessary and sufficient for a particular memory: activating them evokes memory recall; silencing them blocks it.
#Guskjolen, A., Cembrowski, M.S. Engram neurons: Encoding, consolidation, retrieval, and forgetting of memory. Mol Psychiatry (2023).
https://doi.org/10.1038/s41380-023-02137-5
Quote: “Tremendous strides have been made in our understanding of the neurobiological substrates of memory – the so-called memory “engram”. Here, we integrate recent progress in the engram field to illustrate how engram neurons transform across the “lifespan” of a memory — from initial memory encoding, to consolidation and retrieval, and ultimately to forgetting. To do so, we first describe how cell-intrinsic properties shape the initial emergence of the engram at memory encoding. Second, we highlight how these encoding neurons preferentially participate in synaptic- and systems-level consolidation of memory. Third, we describe how these changes during encoding and consolidation guide neural reactivation during retrieval, and facilitate memory recall. Fourth, we describe neurobiological mechanisms of forgetting, and how these mechanisms can counteract engram properties established during memory encoding, consolidation, and retrieval. Motivated by recent experimental results across these four sections, we conclude by proposing some conceptual extensions to the traditional view of the engram, including broadening the view of cell-type participation within engrams and across memory stages. In collection, our review synthesizes general principles of the engram across memory stages, and describes future avenues to further understand the dynamic engram.
[...]
Memory can be defined as an experience-dependent alteration in behavior that persists beyond the environmental stimuli that produced it. Memory is often conceptualized as a multi-staged process that includes encoding, consolidation, retrieval, and forgetting. As such, mechanistically interpreting memory in the brain is facilitated by understanding the neural underpinnings of each of these stages independently, as well as how these neural elements interrelate across stages. In this regard, significant progress has been made in our understanding memory stages at the level of ‘engram neurons’ – that is, neurons that mediate a particular memory across stages [1,2,3,4,5,6,7,8].”
“Fig. 1: Schematic of cellular and synaptic organization across memory stages. The transformation of the engram is depicted for each stage of memory summarized in this review.”
#Ramsey LA, Koya E, van den Oever MC. Editorial: Neuronal ensembles and memory engrams: Cellular and molecular mechanisms. Front Behav Neurosci. 2023
https://pmc.ncbi.nlm.nih.gov/articles/PMC10011704/
Quote: “One of the fundamental questions of neuroscience is how the brain can create, store, and retrieve memories. In the early part of the twentieth century, Karl Lashley attempted to shed light on this question using cortical lesion studies. He concluded that complex behavior relies on both local and distributed storage and retrieval mechanisms in the brain. Lashley (1931) acknowledged that multiple, dispersed brain regions were necessary to enable complex behavior, a principle he referred to as mass action of cerebral function. The search for the mechanistic substrates of memory, what Richard Semon called the “engram” has continued into the present day (Semon, 1921). Later, Hebb (1949). pioneered the idea of neuronal ensembles, which he referred to as “cell assemblies,” or small populations of sparsely distributed neurons active in response to a specific salient stimulus. Another theory put forth by Hebb was the idea that learning occurs via strengthening of synaptic connections between neurons, or synaptic (Hebbian) plasticity. Hebb, Lashley, and their contemporaries had access to a limited toolkit and relied mostly on lesion studies and clinical case studies to test their theories. Fortunately, in the past two decades there has been a renaissance in tools and technology available to identify, characterize, and manipulate neuronal ensembles and engrams (Koya et al., 2009; Kim et al., 2011; Choi et al., 2018; DeNardo et al., 2019; Matos et al., 2019). The advent of these tools has led to an explosion of research that is beginning to uncover the cellular and molecular mechanisms by which memories are encoded and retrieved. This Research Topic contains five papers that further our understanding, both empirically and theoretically, of the cellular and molecular mechanisms within neuronal ensembles that support the engram, using both established methods and cutting-edge technology, as well as incorporating new statistical approaches.”
– Activating any part of the pattern now makes the whole assembly fire – now you are able to relive a moment of the past that is gone forever in the real world.
Human memory recall (especially episodic memory recall, i.e. “mental time travel”) is a highly complex process which emerges from coordinated activity in several different regions of the brain. We are simplifying a lot here, but in a nutshell, recalling a memory relies on the reinstatement and transformation of those neural activity patterns (e.g. cell assemblies, groups of neurons firing together; engrams) originally present when the memory was encoded into the brain.
Some part of the episode/experience, e.g. a smell or sound, can serve as a “cue” to recall the whole memory. This is also called “pattern completion”.
#Miehl, C., Onasch, S., Festa, D. and Gjorgjieva, J. Formation and computational implications of assemblies in neural circuits. J Physiol. (2023)
https://doi.org/10.1113/JP282750
Quote: “The idea of assemblies, or ensembles, as the basic units of cognition has recently replaced the neuron-centric view (Buzsáki, 2010; Eichenbaum, 2018; Huyck & Passmore, 2013; Yuste, 2015). An emergent core assumption from this framework is that each ensemble represents a specific concept or feature, acting as the fundamental unit for memory storage (Neves et al., 2008). One advantage of having strong recurrent connectivity (assembly) compared to only considering correlated rates (ensemble) is stimulus amplification (Peron et al., 2020), thus enabling weaker stimuli to elicit a recognisable response and to increase robustness whereby the malfunction or death of single neurons or synapses will not affect the represented concept. Consequently, an incomplete stimulus can be sufficient to evoke the complete assembly – a phenomenon named pattern completion, which is especially relevant for memory retrieval (Guzman et al., 2016). Moreover, recurrent interactions alone may be sufficient to keep an assembly active after stimulation, thus enabling the brain to decouple intrinsic activity from external stimulation and modify learned concepts independently from specific external inputs (Harris, 2005).”
#Staresina BP, Wimber M. A Neural Chronometry of Memory Recall. Trends Cogn Sci. 2019
https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(19)30235-9
Quote: “
Simple reminders can bring back a host of vivid memories, experimentally epitomised in cued-recall paradigms.
Electrophysiological recordings have elucidated the chronometry with which sensory cues are converted into retrieved memories.
At 500 ms after cue onset, a pattern completion process begins in the hippocampus and triggers the reinstatement of the target memory in the neocortex.
Cortical reinstatement unfolds between 500 and 1500 ms and gives rise to the subjective feeling of recollection.
Reinstatement is governed by intricate temporal dynamics, including the reversal of perceptual processing streams and clocking by theta rhythms.
Episodic memory allows us to mentally travel through time. How does the brain convert a simple reminder cue into a full-blown memory of past events and experiences? In this review, we integrate recent developments in the cognitive neuroscience of human memory retrieval, pinpointing the neural chronometry underlying successful recall. Electrophysiological recordings suggest that sensory cues proceed into the medial temporal lobe within the first 500 ms. At this point, a hippocampal process sets in, geared toward internal pattern completion and coordination of cortical memory reinstatement between 500 and 1500 ms. We further highlight the dynamic principles governing the recall process, which include a reversal of perceptual information flows, temporal compression, and theta clocking.”
#Eichenbaum, H. Prefrontal–hippocampal interactions in episodic memory. Nat Rev Neurosci (2017)
https://doi.org/10.1038/nrn.2017.74
Quote: “The roles of the hippocampus and prefrontal cortex (PFC) in memory processing — individually or in concert — are a major topic of interest in memory research. These brain areas have distinct and complementary roles in episodic memory, and their interactions are crucial for learning and remembering events. Considerable evidence indicates that the PFC and hippocampus become coupled via oscillatory synchrony that reflects bidirectional flow of information. Furthermore, newer studies have revealed specific mechanisms whereby neural representations in the PFC and hippocampus are mediated through direct connections or through intermediary regions. These findings suggest a model of how the hippocampus and PFC, along with their intermediaries, operate as a system that uses the current context of experience to retrieve relevant memories.”
#Franziska R Richter, Rose A Cooper, Paul M Bays, Jon S Simons. Distinct neural mechanisms underlie the success, precision, and vividness of episodic memory. eLife (2016)
https://elifesciences.org/articles/18260
Quote: “A network of brain regions have been linked with episodic memory retrieval, but limited progress has been made in identifying the contributions of distinct parts of the network. Here, we utilized continuous measures of retrieval to dissociate three components of episodic memory: retrieval success, precision, and vividness. In the fMRI scanner, participants encoded objects that varied continuously on three features: color, orientation, and location. Participants’ memory was tested by having them recreate the appearance of the object features using a continuous dial, and continuous vividness judgments were recorded. Retrieval success, precision, and vividness were dissociable both behaviorally and neurally: successful versus unsuccessful retrieval was associated with hippocampal activity, retrieval precision scaled with activity in the angular gyrus, and vividness judgments tracked activity in the precuneus. The ability to dissociate these components of episodic memory reveals the benefit afforded by measuring memory on a continuous scale, allowing functional parcellation of the retrieval network.”
– But this new memory is very fragile and still pretty temporary. Your hippocampus holds its blueprint, but without reinforcement, the assembly will fade and the synapses will get weakened again. This is why you forget most moments of your life, why you don’t remember how your coffee tasted 43 weeks ago on a Monday.
For a memory to last a long time, it must not just be encoded into the brain once, but also then “consolidated”. This means that the memory is turned from soft, fragile “traces” in the brain (short-term memory) to a long-lasting, stable memory (long-term memory). Memory consolidation is a highly complex, multi-phase process and is vulnerable to disruption, especially in the first hours to days.
Importantly, not all memories get consolidated. Quite the opposite: the brain is highly selective in what gets consolidated. It acts as a big filter for what is “worth remembering”. But it is also not binary, so there is no hard line between “will definitely remember” and “will definitely forget”. It is a gradual change of likelihood of remembering vs. forgetting.
In the example of getting a random coffee one day, the brain will likely not engage in (much) consolidation, meaning the moment will be encoded once but then will likely fade away rather quickly. Without being put into long-term storage, without being consolidated, it will be difficult or impossible to recall this exact moment of getting coffee many weeks/months later.
Consolidation happens in two nested levels at different timescales: synaptic consolidation (hours) and systems consolidation (days to weeks). There are several biological mechanisms underlying consolidation, and many things are also not yet well understood. One process which is thought to be important for the transformation into long-term memory is the strengthening of the synapses within the neural representation of the memory, also called an “engram”. When this doesn’t happen, the synapses can get weakened instead.
But we are also simplifying a lot here. Learning events trigger a cascade of cellular and molecular processes that must happen for a memory to be stored. If this does not happen, it only lasts as working memory or short-term memory (for minutes-hours). In this video we only give a general overview of the broad strokes process.
#Sridhar S, Khamaj A and Asthana MK. Cognitive neuroscience perspective on memory: overview and summary. Front. Hum. Neurosci. (2023)
https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2023.1217093/full
Quote: “Memory is an essential cognitive function that permits individuals to acquire, retain, and recover data that defines a person’s identity (Zlotnik and Vansintjan, 2019). Memory is a multifaceted cognitive process that involves different stages: encoding, consolidation, recovery, and reconsolidation. Encoding involves acquiring and processing information that is transformed into a neuronal representation suitable for storage (Liu et al., 2021; Panzeri et al., 2023). The information can be acquired through various channels, such as visual, auditory, olfactory, or tactile inputs. The acquired sensory stimuli are converted into a format the brain can process and retain. Different factors such as attention, emotional significance, and repetition can influence the encoding process and determine the strength and durability of the resulting memory (Squire et al., 2004; Lee et al., 2016; Serences, 2016).
Consolidation includes the stabilization and integration of memory into long-term storage to increase resistance to interference and decay (Goedert and Willingham, 2002). This process creates enduring structural modification in the brain and thereby has consequential effects on the function by reorganizing and strengthening neural connections. Diverse sources like sleep and stress and the release of neurotransmitters can influence memory consolidation. Many researchers have noted the importance of sleep due to its critical role in enabling a smooth transition of information from transient repositories into more stable engrams (memory traces) (McGaugh, 2000; Clawson et al., 2021; Rakowska et al., 2022).
[...]
For an event to be remembered, it must form physical connections between neurons in the brain, which creates a “memory trace.” This memory trace can then be stored as long-term memory (Langille and Brown, 2018). The formation of a memory engram is an intricate process requiring neuronal depolarization and the influx of intracellular calcium (Mank and Griesbeck, 2008; Josselyn et al., 2015; Xu et al., 2017). This initiation leads to a cascade involving protein transcription, structural and functional changes in neural networks, and stabilization during the quiescence period, followed by complete consolidation for its success. Interference from new learning events or disruption caused due to inhibition can abort this cycle leading to incomplete consolidation (Josselyn et al., 2015).
Cyclic-AMP response element binding protein (CREB) has been identified as an essential transcription factor for memory formation (Orsini and Maren, 2012). It regulates the expression of PRPs and enhances neuronal excitability and plasticity, resulting in changes to the structure of cells, including the growth of dendritic spines and new synaptic connections. Blockage or enhancement of CREB in certain areas can affect subsequent consolidation at a systems level–decreasing it prevents this from occurring, while aiding its presence allows even weak learning conditions to produce successful memory formation (Orsini and Maren, 2012; Kandel et al., 2014).
Strengthening weakly encoded memories through the synaptic tagging and capture hypothesis may play an essential role in cellular consolidation. Retroactive memory enhancement has also been demonstrated in human studies, mainly when items are initially encoded with low strength but later paired with shock after consolidation (Dunsmoor et al., 2015). The synaptic tagging and capture theory (STC) and its extension, the behavioral tagging hypothesis (BT), have both been used to explain synaptic specificity and the persistence of plasticity (Moncada et al., 2015). STC proposed that electrophysiological activity can induce long-term changes in synapses, while BT postulates similar effects of behaviorally relevant neuronal events on learning and memory models. This hypothesis proposes that memory consolidation relies on combining two distinct processes: setting a “learning tag” and synthesizing plasticity-related proteins (De novo protein synthesis, increased CREB levels, and substantial inputs to nearby synapses) at those tagged sites. BT explains how it is possible for event episodes with low-strength inputs or engagements can be converted into lasting memories (Lynch, 2004; Moncada et al., 2015). Similarly, the emotional tagging hypothesis posits that the activation of the amygdala in emotionally arousing events helps to mark experiences as necessary, thus enhancing synaptic plasticity and facilitating transformation from transient into more permanent forms for encoding long-term memories (Richter-Levin and Akirav, 2003; Zhu et al., 2022).
[...]
Cellular consolidation, the protein synthesis-dependent processes observed in rodents that may underlie memory formation and stabilization, has been challenging to characterize in humans due to the limited ability to study it directly (Bermudez-Rattoni, 2010). Additionally, multi-trial learning protocols commonly used within human tests as opposed to single-trial experiments conducted with non-human subjects suggest there could be interference from subsequent information that impedes individual memories from being consolidated reliably. This raises important questions regarding how individuals can still form strong and long-lasting memories when exposed to frequent stimuli outside controlled laboratory conditions. Although this phenomenon remains undiscovered by science, it is of utmost significance for gaining a deeper understanding of our neural capacities (Genzel and Wixted, 2017).
The establishment of distributed memory traces requires a narrow temporal window following the initial encoding process, during which cellular consolidation occurs (Nader and Hardt, 2009). Once this period ends and consolidation has been completed, further protein synthesis inhibition or pharmacological disruption will be less effective at altering pre-existing memories and interfering with new learning due to the stabilization of the trace in its new neuronal network connections (Nader and Hardt, 2009). Thus, systems consolidation appears critical for the long-term maintenance of memory within broader brain networks over extended periods after their formation (Bermudez-Rattoni, 2010).
[...]
Information is initially stored in both the hippocampus and neocortex (Dudai et al., 2015). The hippocampus subsequently guides a gradual process of reorganization and stabilization whereby information present within the neocortex becomes autonomous from that in the hippocampal store. Scholars have termed this phenomenon “standard memory consolidation model” or “system consolidation” (Squire et al., 2015).
The Standard Model suggests that information acquired during learning is simultaneously stored in both the hippocampus and multiple cortical modules. Subsequently, it posits that over a period of time which may range from weeks to months or longer, the hippocampal formation directs an integration process by which these various elements become enclosed into single unified structures within the cortex (Gilboa and Moscovitch, 2021; Howard et al., 2022). These newly learned memories are then assimilated into existing networks without interference or compression when necessary (Frankland and Bontempi, 2005). It is important to note that memory engrams already exist within cortical networks during encoding. They only need strengthening through links enabled by hippocampal assistance-overtime allowing remote memory storage without reliance on the latter structure. Data appears consistent across studies indicating that both AMPA-and NMDA receptor-dependent “tagging” processes occurring within the cortex are essential components of progressive rewiring, thus enabling longer-term retention (Takeuchi et al., 2014; Takehara-Nishiuchi, 2020).”
#Bin Ibrahim, M.Z., Benoy, A. and Sajikumar, S. Long-term plasticity in the hippocampus: maintaining within and ‘tagging’ between synapses. FEBS J. (2022)
https://doi.org/10.1111/febs.16065
Quote: “Synapses between neurons are malleable biochemical structures, strengthening and diminishing over time dependent on the type of information they receive. This phenomenon known as synaptic plasticity underlies learning and memory, and its different forms, long-term potentiation (LTP) and long-term depression (LTD), perform varied cognitive roles in reinforcement, relearning and associating memories. Moreover, both LTP and LTD can exist in an early transient form (early-LTP/LTD) or a late persistent form (late-LTP/LTD), which are triggered by different induction protocols, and also differ in their dependence on protein synthesis and the involvement of key molecular players. Beyond homosynaptic modifications, synapses can also interact with one another. This is encapsulated in the synaptic tagging and capture hypothesis (STC), where synapses expressing early-LTP/LTD present a ‘tag’ that can capture the protein synthesis products generated during a temporally proximal late-LTP/LTD induction. This ‘tagging’ phenomenon forms the framework of synaptic interactions in various conditions and accounts for the cellular basis of the time-dependent associativity of short-lasting and long-lasting memories. All these synaptic modifications take place under controlled neuronal conditions, regulated by subcellular elements such as epigenetic regulation, proteasomal degradation and neuromodulatory signals. Here, we review current understanding of the different forms of synaptic plasticity and its regulatory mechanisms in the hippocampus, a brain region critical for memory formation. We also discuss expression of plasticity in hippocampal CA2 area, a long-overlooked narrow hippocampal subfield and the behavioural correlate of STC. Lastly, we put forth perspectives for an integrated view of memory representation in synapses.”
– One of them is novelty. If you walk to the bus, listening to some mildly interesting podcast as usual, this assembly’s signal is too weak. This moment in time will be lost, like tears in rain. But if one day you see a crow and a squirrel fighting over a nut, only for a mouse to steal it – the assembly will fire strongly just for the novelty of it all.
A novel stimulus can increase the likelihood of forming a long-lasting memory. But the relationship between stimulus novelty and memory formation is not linear and highly nuanced: it depends on the type of stimulus and the context.
In addition to the sense of novelty, your brain might also experience emotions such as joy in observing the crow/squirrel scenario. Emotions can also increase the chances of memory formation. We cover this scenario later in the video, when we talk about the importance of emotions for memory.
#Reichardt, R., Polner, B. & Simor, P. The graded novelty encoding task: Novelty gradually improves recognition of visual stimuli under incidental learning conditions. Behav Res (2023)
https://doi.org/10.3758/s13428-022-01891-8
Quote: “Memory is highly selective, and the rules governing the entry of information into memory systems are a central concern of memory research (Baddeley et al., 2020; Eichenbaum, 2017). Among the many variables influencing memory formation, novelty is probably the most intimately connected with memory, since it is defined in terms of what is already represented in memory (Barto et al., 2013). Several researchers have put forth theories on why and how novelty plays a central role in memory formation (Barto et al., 2013; Duszkiewicz et al., 2019; Frank & Kafkas, 2021; Press et al., 2020; Quent et al., 2021; Reichardt et al., 2020; Schomaker & Meeter, 2015; Van Kesteren et al., 2012).
Early on, Tulving and colleagues formulated a theory on the beneficial effect of novelty on memory: they noticed that several brain regions, among them the hippocampus, responded differently to novel than to familiar stimuli (Tulving et al., 1996; Tulving & Kroll, 1995). This inspired the novelty/encoding hypothesis (Tulving & Kroll, 1995), which states that the hippocampal activation elicited by novel stimuli is due to a memory encoding process instantiated by novelty detection. According to this classic theory, memory performance should be better for novel items than for familiar ones due to this increased hippocampal activation.
However, the novelty/encoding hypothesis is too simplistic to describe the full breadth of findings related to the memory effects of novelty (Barto et al., 2013; Duszkiewicz et al., 2019; Kafkas & Montaldi, 2018; Press et al., 2020; Reichardt et al., 2020). More specifically, familiar stimuli are sometimes better remembered than novel ones (Poppenk et al., 2010), and neural responses to novel stimuli reduce dramatically over time (Murty et al., 2013; Nieuwenhuis et al., 2011). Furthermore, challenges emerged at the conceptual level as well, in that novelty can be assorted into different categories which may elicit different memory processes (Duszkiewicz et al., 2019; Frank & Kafkas, 2021; Quent et al., 2021; Reichardt et al., 2020). For example, Berlyne, in his pioneering studies of novelty and its effects on cognition, distinguished between complete and relative novelty (Berlyne, 1960). Complete novelty is attributed to a stimulus that the observer has never encountered before, while a stimulus with relative novelty may only be novel due to the unfamiliar combination of well-known features. More recently, authors differentiated between absolute and contextual novelty, although these categories are essentially the same as complete and relative novelty (Kafkas & Montaldi, 2018). Since Berlyne, many researchers have emphasized that novelty should be understood as a continuum which is theoretically crucial in that it allows us to study the dose–response effects of novelty on memory (Frank & Kafkas, 2021; Quent et al., 2021; Reichardt et al., 2020). And indeed, recent cognitive neuroscientific studies yielded convergent evidence showing that the degree of novelty has a major influence on memory formation (Duszkiewicz et al., 2019; Reichardt et al., 2020; Van Kesteren et al., 2012).
[...]
Here, using a novel experimental paradigm, we demonstrate that novelty improves subsequent recognition memory in a dose-dependent manner. We validated our continuous novelty manipulation by showing that the more different a stimulus is from familiarized stimuli, the more likely that it will be distinguished from the familiars during the study phase. Overall, the findings fit with theories of the beneficial effect of novelty on memory and can be interpreted in a predictive processing framework, opening avenues for the application of our paradigm in future neuroimaging and computational experiments.”
#Frank D, Kafkas A. Expectation-driven novelty effects in episodic memory. Neurobiol Learn Mem. 2021
https://pubmed.ncbi.nlm.nih.gov/34048914/
Quote: “Novel and unexpected stimuli are often prioritised in memory, given their inherent salience. Nevertheless, not all forms of novelty show such an enhancement effect. Here, we discuss the role expectation plays in modulating the way novelty affects memory processes, circuits, and subsequent performance. We first review independent effects of expectation on memory, and then consider how different types of novelty are characterised by expectation. We argue that different types of novelty defined by expectation implicate differential neurotransmission in memory formation brain regions and may also result in the creation of different types of memory. Contextual novelty, which is unexpected by definition, is often associated with better recollection, supported by dopaminergic-hippocampal interactions. On the other hand, expected stimulus novelty is supported by engagement of medial temporal cortices, as well as the hippocampus, through cholinergic modulation. Furthermore, when expected stimulus novelty results in enhanced memory, it is predominantly driven by familiarity. The literature reviewed here highlights the complexity of novelty-sensitive memory systems, the distinction between types of novelty, and how they are differentially affected by expectancy.”
#Duszkiewicz AJ, McNamara CG, Takeuchi T, Genzel L. Novelty and Dopaminergic Modulation of Memory Persistence: A Tale of Two Systems. Trends Neurosci. 2019
https://pmc.ncbi.nlm.nih.gov/articles/PMC6352318/
Quote: “Adaptation to the ever-changing world is critical for survival, and our brains are particularly tuned to remember events that differ from previous experiences. Novel experiences induce dopamine release in the hippocampus, a process which promotes memory persistence. While axons from the ventral tegmental area (VTA) were generally thought to be the exclusive source of hippocampal dopamine, recent studies have demonstrated that noradrenergic neurons in the locus coeruleus (LC) corelease noradrenaline and dopamine in the hippocampus and that their dopamine release boosts memory retention as well. In this opinion article, we propose that the projections originating from the VTA and the LC belong to two distinct systems that enhance memory of novel events. Novel experiences that share some commonality with past ones (‘common novelty’) activate the VTA and promote semantic memory formation via systems memory consolidation. By contrast, experiences that bear only a minimal relationship to past experiences (‘distinct novelty’) activate the LC to trigger strong initial memory consolidation in the hippocampus, resulting in vivid and long-lasting episodic memories.”
#Yamasaki, Miwako, Takeuchi, Tomonori, Locus Coeruleus and Dopamine-Dependent Memory Consolidation, Neural Plasticity. 2017
https://doi.org/10.1155/2017/8602690
Quote: “Many people have vivid memories of the first dinner date with their partner, including details like the name of the restaurant and the food they had. In contrast, it is very difficult to remember what you had for dinner a few weeks ago. Most everyday memories, including episodic-like memories that we may form automatically in the hippocampus (HPC) [1–3], are forgotten, whereas some of them are retained for a long time by a memory stabilization process (initial memory consolidation). Initial selective retention occurs when something novel or salient happens shortly before or after the time of memory encoding, as in “flashbulb memory” [4, 5]. Unexpected novel events create a “halo” of enhanced memory, triggering an initial memory consolidation which extends not only forwards but also backwards in time, boosting retention of trivial memories that would normally be forgotten. Thus, initial consolidation serves as the “gate” to long-term memory, so that only a subset of information is retained for long enough to be subject to stabilization in the neocortex via a complementary process of “systems memory consolidation” [6, 7].”
– Another one is to reactivate the memory over and over after it is formed. Thinking about the animal fight all day and telling everyone about the strange event will etch it deeper into your brain. Similar to doing loads of repetition when you try to learn something.
When a memory is recalled over and over again, it can be maintained for longer periods of time. This is basically the “learning effect” when e.g. studying for a test.
#W. Yu, A. Zadbood, A.J.H. Chanales, & L. Davachi, Repetition dynamically and rapidly increases cortical, but not hippocampal, offline reactivation, Proc. Natl. Acad. Sci. U.S.A. (2024)
https://www.pnas.org/doi/10.1073/pnas.2405929121
Quote: “The distribution of memories across hippocampal-cortical networks is a hallmark of memory consolidation. While repeated study has been shown to improve retention, the mechanisms supporting these effects remain unknown. Here, we show that repetition increases post-encoding offline reactivation in the cortex and enhances the coordinated offline reactivation between the hippocampus and cortex, providing important evidence that memory consolidation in hippocampal-cortical networks can be robustly accelerated through repeated learning. Further, we demonstrate that offline reactivation in both the hippocampus and cortex explains variance in the memory outcomes of once-encoded memories. Therefore, while prioritizing repeated memories, offline reactivation may also compensate for inadequate encoding to achieve balanced consolidation across memories.”
#L. Himmer et al. Rehearsal initiates systems memory consolidation, sleep makes it last. Sci. Adv. (2019)
https://www.science.org/doi/10.1126/sciadv.aav1695
Quote: “After encoding, memories undergo a transitional process termed systems memory consolidation. It allows fast acquisition of new information by the hippocampus, as well as stable storage in neocortical long-term networks, where memory is protected from interference. Whereas this process is generally thought to occur slowly over time and sleep, we recently found a rapid memory systems transition from hippocampus to posterior parietal cortex (PPC) that occurs over repeated rehearsal within one study session. Here, we use fMRI to demonstrate that this transition is stabilized over sleep, whereas wakefulness leads to a reset to naïve responses, such as observed during early encoding. The role of sleep therefore seems to go beyond providing additional rehearsal through memory trace reactivation, as previously thought. We conclude that repeated study induces systems consolidation, while sleep ensures that these transformations become stable and long lasting. Thus, sleep and repeated rehearsal jointly contribute to long-term memory consolidation.”
#Zhan L, Guo D, Chen G, Yang J. Effects of Repetition Learning on Associative Recognition Over Time: Role of the Hippocampus and Prefrontal Cortex. Front Hum Neurosci. 2018
https://pmc.ncbi.nlm.nih.gov/articles/PMC6050388/
Quote: “When stimuli are learned by repetition, they are remembered better and retained for a longer time. However, current findings are lacking as to whether the medial temporal lobe (MTL) and cortical regions are involved in the learning effect when subjects retrieve associative memory, and whether their activations differentially change over time due to learning experience. To address these issues, we designed an fMRI experiment in which face-scene pairs were learned once (L1) or six times (L6). Subjects learned the pairs at four retention intervals, 30-min, 1-day, 1-week and 1-month, after which they finished an associative recognition task in the scanner. The results showed that compared to learning once, learning six times led to stronger activation in the hippocampus, but weaker activation in the perirhinal cortex (PRC) as well as anterior ventrolateral prefrontal cortex (vLPFC). In addition, the hippocampal activation was positively correlated with that of the parahippocampal place area (PPA) and negatively correlated with that of the vLPFC when the L6 group was compared to the L1 group. The hippocampal activation decreased over time after L1 but remained stable after L6. These results clarified how the hippocampus and cortical regions interacted to support associative memory after different learning experiences.”
– Emotions are really strong mechanisms to guide our behavior that evolved hundreds of millions of years ago. They motivate you to avoid danger, seek out food and reproduce. You experience this as something feeling good or bad. Whenever you feel something strongly, your ancient brain decides that whatever is going on is important for your survival, regardless if it's wrong or not.
What exactly emotions are in a biological sense, whether animals have them (and to what degree), or if they only exist in the human context, has been a long debated topic in biology, psychology and philosophy. That being said, basic brain states important for survival (e.g. pleasure from eating food) are likely ancient, dating back at least to the evolution of mammals. However, this is an area of active research and many things are not yet well understood.
#Anderson DJ, Adolphs R. A framework for studying emotions across species. Cell. 2014
https://pmc.ncbi.nlm.nih.gov/articles/PMC4098837/
Quote: “Since the 19th century, there has been disagreement over the fundamental question of whether “emotions” are cause or consequence of their associated behaviors. This question of causation is most directly addressable in genetically tractable model organisms, including invertebrates such as Drosophila. Yet there is ongoing debate about whether such species even have “emotions,” as emotions are typically defined with reference to human behavior and neuroanatomy. Here, we argue that emotional behaviors are a class of behaviors that express internal emotion states. These emotion states exhibit certain general functional and adaptive properties that apply across any specific human emotions like fear or anger, as well as across phylogeny. These general properties, which can be thought of as “emotion primitives,” can be modeled and studied in evolutionarily distant model organisms, allowing functional dissection of their mechanistic bases and tests of their causal relationships to behavior. More generally, our approach not only aims at better integration of such studies in model organisms with studies of emotion in humans, but also suggests a revision of how emotion should be operationalized within psychology and psychiatry.”
#LeDoux J. Rethinking the emotional brain. Neuron. 2012
https://pmc.ncbi.nlm.nih.gov/articles/PMC3625946/
Quote: “I propose a reconceptualization of key phenomena important in the study of emotion—those phenomena that reflect functions and circuits related to survival, and that are shared by humans and other animals. The approach shifts the focus from questions about whether emotions that humans consciously feel are also present in other animals, and toward questions about the extent to which circuits and corresponding functions that are present in other animals (survival circuits and functions) are also present in humans. Survival circuit functions are not causally related to emotional feelings but obviously contribute to these, at least indirectly. The survival circuit concept integrates ideas about emotion, motivation, reinforcement, and arousal in the effort to understand how organisms survive and thrive by detecting and responding to challenges and opportunities in daily life.
[...]
In this Perspective I, therefore, describe a way of conceiving phenomena important to the study of emotion, but with minimal recourse to the terms emotion or feelings. The focus is instead on circuits that instantiate functions that allow organisms to survive and thrive by detecting and responding to challenges and opportunities. Included, at a minimum, are circuits involved in defense, maintenance of energy and nutritional supplies, fluid balance, thermoregulation, and reproduction. These survival circuits and their adaptive functions are conserved to a significant degree in across mammalian species, including humans. While there are species-specific aspects of these functions, there are also core components of these functions that are shared by all mammals.”
#Armstrong, D.M., Jones, J.K., Wilson, D.E. "mammal." Encyclopedia Britannica, December 22, 2025. Retrieved January 2026
https://www.britannica.com/animal/mammal
“The evolution of the mammalian condition. Mammals were derived in the Triassic Period (about 252 million to 201 million years ago) from members of the reptilian order Therapsida.”
– So many of your strongest memories have strong emotional flavors.
#McGaugh, J. L. Emotional arousal regulation of memory consolidation. Current Opinion in Behavioral Sciences. (2018)
https://doi.org/10.1016/j.cobeha.2017.10.003
Quote: “•Emotional activation before, during or shortly after learning enhances memory.
•Memory enhancement is induced by stress hormone activation of the amygdala.
•The amygdala activates brain regions that process different forms of memory.
Recent findings provide increased understanding of how emotional arousal creates lasting memories. The findings are consistent with those of prior studies suggesting that the enhancement assessed in human subjects results from activation of adrenergic and glucocorticoid stress hormones. Additionally, fMRI imaging findings indicate that the enhancement is influenced by activation of the amygdala and its subsequent influences on other brain systems. Findings of recent animal studies using posttraining noradrenergic or optogenetic activation of the amygdala provide extensive evidence that the basolateral amygdala modulates memory consolidation by influencing neuroplasticity in downstream brain systems involved in processing different forms of memory. Activation of these systems helps to insure that emotionally significant experiences are well remembered.”
#J.L. McGaugh, Making lasting memories: Remembering the significant, Proc. Natl. Acad. Sci. U.S.A. (2013)
https://doi.org/10.1073/pnas.1301209110
Quote: “There is also extensive evidence that experiences that are emotionally arousing are well-remembered (7–10). Experiences of unpleasant occasions, such as an automobile accident, a mugging, or learning about the death of a loved one, are remembered better than those experiences of a routine day (8, 11–18). Memories of pleasant occasions, such as birthdays, holidays, and weddings, are also well-retained. The strength of memories of events varies with the emotional significance of the events. The memories of individuals who were close to San Francisco at the time of the 1989 San Francisco earthquake had better memories of the earthquake months later compared with individuals in Atlanta, Georgia (16). Three years after the terrorist attack on September 11, 2001, individuals who were in downtown Manhattan at the time of the attack had more detailed memories of the attack compared with individuals who were in midtown Manhattan, several miles from the attack (18).”
– Strong activation, repetition and emotions do the same thing to your memory: We said before that chemicals made the neurons able to change – well now they change drastically. Like wax warming up and melting together, the gears of the assembly grow new teeth to fit more tightly – neurons grow more synapses and fire together even better. They become closer and more solid.
For a memory to last a long time, it must not just be encoded into the brain once, but also then “consolidated”. This means that the memory is turned from soft, fragile traces in the brain (short-term memory) to a long-lasting, stable memory (long-term memory). Memory consolidation is a highly complex, multi-phase process and is vulnerable to disruption, especially in the first hours to days.
It is not a guarantee or a perfect predictor, but strong neural activation (through e.g. novelty or heightened attention), repetition and emotions can strengthen the consolidation process, making it more likely that an event or detail will be remembered.
Consolidation happens in two nested levels at different timescales: synaptic consolidation (hours) and systems consolidation (days to weeks). There are several biological mechanisms underlying consolidation, and many things are not yet well understood. One process which is thought to be important for the transformation into long-term memory is the strengthening of the synapses within the neural representation of the memory, also called an “engram”.
#Sridhar S, Khamaj A and Asthana MK. Cognitive neuroscience perspective on memory: overview and summary. Front. Hum. Neurosci. (2023)
https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2023.1217093/full
Quote: “Memory is an essential cognitive function that permits individuals to acquire, retain, and recover data that defines a person’s identity (Zlotnik and Vansintjan, 2019). Memory is a multifaceted cognitive process that involves different stages: encoding, consolidation, recovery, and reconsolidation. Encoding involves acquiring and processing information that is transformed into a neuronal representation suitable for storage (Liu et al., 2021; Panzeri et al., 2023). The information can be acquired through various channels, such as visual, auditory, olfactory, or tactile inputs. The acquired sensory stimuli are converted into a format the brain can process and retain. Different factors such as attention, emotional significance, and repetition can influence the encoding process and determine the strength and durability of the resulting memory (Squire et al., 2004; Lee et al., 2016; Serences, 2016).
Consolidation includes the stabilization and integration of memory into long-term storage to increase resistance to interference and decay (Goedert and Willingham, 2002). This process creates enduring structural modification in the brain and thereby has consequential effects on the function by reorganizing and strengthening neural connections. Diverse sources like sleep and stress and the release of neurotransmitters can influence memory consolidation. Many researchers have noted the importance of sleep due to its critical role in enabling a smooth transition of information from transient repositories into more stable engrams (memory traces) (McGaugh, 2000; Clawson et al., 2021; Rakowska et al., 2022).
[...]
For an event to be remembered, it must form physical connections between neurons in the brain, which creates a “memory trace.” This memory trace can then be stored as long-term memory (Langille and Brown, 2018). The formation of a memory engram is an intricate process requiring neuronal depolarization and the influx of intracellular calcium (Mank and Griesbeck, 2008; Josselyn et al., 2015; Xu et al., 2017). This initiation leads to a cascade involving protein transcription, structural and functional changes in neural networks, and stabilization during the quiescence period, followed by complete consolidation for its success. Interference from new learning events or disruption caused due to inhibition can abort this cycle leading to incomplete consolidation (Josselyn et al., 2015).
Cyclic-AMP response element binding protein (CREB) has been identified as an essential transcription factor for memory formation (Orsini and Maren, 2012). It regulates the expression of PRPs and enhances neuronal excitability and plasticity, resulting in changes to the structure of cells, including the growth of dendritic spines and new synaptic connections. Blockage or enhancement of CREB in certain areas can affect subsequent consolidation at a systems level–decreasing it prevents this from occurring, while aiding its presence allows even weak learning conditions to produce successful memory formation (Orsini and Maren, 2012; Kandel et al., 2014).
Strengthening weakly encoded memories through the synaptic tagging and capture hypothesis may play an essential role in cellular consolidation. Retroactive memory enhancement has also been demonstrated in human studies, mainly when items are initially encoded with low strength but later paired with shock after consolidation (Dunsmoor et al., 2015). The synaptic tagging and capture theory (STC) and its extension, the behavioral tagging hypothesis (BT), have both been used to explain synaptic specificity and the persistence of plasticity (Moncada et al., 2015). STC proposed that electrophysiological activity can induce long-term changes in synapses, while BT postulates similar effects of behaviorally relevant neuronal events on learning and memory models. This hypothesis proposes that memory consolidation relies on combining two distinct processes: setting a “learning tag” and synthesizing plasticity-related proteins (De novo protein synthesis, increased CREB levels, and substantial inputs to nearby synapses) at those tagged sites. BT explains how it is possible for event episodes with low-strength inputs or engagements can be converted into lasting memories (Lynch, 2004; Moncada et al., 2015). Similarly, the emotional tagging hypothesis posits that the activation of the amygdala in emotionally arousing events helps to mark experiences as necessary, thus enhancing synaptic plasticity and facilitating transformation from transient into more permanent forms for encoding long-term memories (Richter-Levin and Akirav, 2003; Zhu et al., 2022).
[...]
Cellular consolidation, the protein synthesis-dependent processes observed in rodents that may underlie memory formation and stabilization, has been challenging to characterize in humans due to the limited ability to study it directly (Bermudez-Rattoni, 2010). Additionally, multi-trial learning protocols commonly used within human tests as opposed to single-trial experiments conducted with non-human subjects suggest there could be interference from subsequent information that impedes individual memories from being consolidated reliably. This raises important questions regarding how individuals can still form strong and long-lasting memories when exposed to frequent stimuli outside controlled laboratory conditions. Although this phenomenon remains undiscovered by science, it is of utmost significance for gaining a deeper understanding of our neural capacities (Genzel and Wixted, 2017).
The establishment of distributed memory traces requires a narrow temporal window following the initial encoding process, during which cellular consolidation occurs (Nader and Hardt, 2009). Once this period ends and consolidation has been completed, further protein synthesis inhibition or pharmacological disruption will be less effective at altering pre-existing memories and interfering with new learning due to the stabilization of the trace in its new neuronal network connections (Nader and Hardt, 2009). Thus, systems consolidation appears critical for the long-term maintenance of memory within broader brain networks over extended periods after their formation (Bermudez-Rattoni, 2010).
[...]
Information is initially stored in both the hippocampus and neocortex (Dudai et al., 2015). The hippocampus subsequently guides a gradual process of reorganization and stabilization whereby information present within the neocortex becomes autonomous from that in the hippocampal store. Scholars have termed this phenomenon “standard memory consolidation model” or “system consolidation” (Squire et al., 2015).
The Standard Model suggests that information acquired during learning is simultaneously stored in both the hippocampus and multiple cortical modules. Subsequently, it posits that over a period of time which may range from weeks to months or longer, the hippocampal formation directs an integration process by which these various elements become enclosed into single unified structures within the cortex (Gilboa and Moscovitch, 2021; Howard et al., 2022). These newly learned memories are then assimilated into existing networks without interference or compression when necessary (Frankland and Bontempi, 2005). It is important to note that memory engrams already exist within cortical networks during encoding. They only need strengthening through links enabled by hippocampal assistance-overtime allowing remote memory storage without reliance on the latter structure. Data appears consistent across studies indicating that both AMPA-and NMDA receptor-dependent “tagging” processes occurring within the cortex are essential components of progressive rewiring, thus enabling longer-term retention (Takeuchi et al., 2014; Takehara-Nishiuchi, 2020).”
#Bin Ibrahim, M.Z., Benoy, A. and Sajikumar, S. Long-term plasticity in the hippocampus: maintaining within and ‘tagging’ between synapses. FEBS J. (2022)
https://doi.org/10.1111/febs.16065
Quote: “Synapses between neurons are malleable biochemical structures, strengthening and diminishing over time dependent on the type of information they receive. This phenomenon known as synaptic plasticity underlies learning and memory, and its different forms, long-term potentiation (LTP) and long-term depression (LTD), perform varied cognitive roles in reinforcement, relearning and associating memories. Moreover, both LTP and LTD can exist in an early transient form (early-LTP/LTD) or a late persistent form (late-LTP/LTD), which are triggered by different induction protocols, and also differ in their dependence on protein synthesis and the involvement of key molecular players. Beyond homosynaptic modifications, synapses can also interact with one another. This is encapsulated in the synaptic tagging and capture hypothesis (STC), where synapses expressing early-LTP/LTD present a ‘tag’ that can capture the protein synthesis products generated during a temporally proximal late-LTP/LTD induction. This ‘tagging’ phenomenon forms the framework of synaptic interactions in various conditions and accounts for the cellular basis of the time-dependent associativity of short-lasting and long-lasting memories. All these synaptic modifications take place under controlled neuronal conditions, regulated by subcellular elements such as epigenetic regulation, proteasomal degradation and neuromodulatory signals. Here, we review current understanding of the different forms of synaptic plasticity and its regulatory mechanisms in the hippocampus, a brain region critical for memory formation. We also discuss expression of plasticity in hippocampal CA2 area, a long-overlooked narrow hippocampal subfield and the behavioural correlate of STC. Lastly, we put forth perspectives for an integrated view of memory representation in synapses.”
– A lot of this happens while you sleep. Your hippocampus replays the assembly over and over, making it more solid and easier to retrieve – which also means that if you don’t sleep enough, you literally forget more of your life. Think about that when you have to study for a test next time – without proper sleep, you’ll be wasting your time.
Research has shown that sleep deprivation negatively affects human memory. For example, sleep deprivation increases the rates of forgetting and decreases the quality/accuracy of memories. This is especially evident in formal learning/educational contexts, where a lot of new information is memorized in a short period of time. Lack of sleep during those periods can severely decrease the amount of successfully retained information.
Note that in the video we are using “forgetting” here in the colloquial sense of the word, which can mean “forming a memory but then losing it over time” but also “never forming a long-term memory in the first place”. In neurobiology these two phenomena are treated as different, because they are distinct processes in the brain.
#Crowley R, Alderman E, Javadi AH, Tamminen J. A systematic and meta-analytic review of the impact of sleep restriction on memory formation. Neurosci Biobehav Rev. 2024
https://www.sciencedirect.com/science/article/pii/S0149763424003981?via%3Dihub
Quote: “Modern life causes a quarter of adults and half of teenagers to sleep for less than is recommended (Kocevska et al., 2021). Given well-documented benefits of sleep on memory, we must understand the cognitive costs of short sleep. We analysed 125 sleep restriction effect sizes from 39 reports involving 1234 participants. Restricting sleep (3–6.5 hours) compared to normal sleep (7–11 hours) negatively affects memory formation with a small effect size (Hedges’ g = 0.29, 95 % CI = [0.13, 0.44]). We detected no evidence for publication bias. When sleep restriction effect sizes were compared with 185 sleep deprivation effect sizes (Newbury et al., 2021) no statistically significant difference was found, suggesting that missing some sleep has similar consequences for memory as not sleeping at all. When the analysis was restricted to pre-encoding, rather than post-encoding, sleep loss, sleep deprivation was associated with larger memory impairment than restriction. Our findings are best accounted for by the sequential hypothesis which emphasises complementary roles of slow-wave sleep and REM sleep for memory.”
#Bolsius YG, Heckman PRA, Paraciani C, Wilhelm S, Raven F, Meijer EL, Kas MJH, Ramirez S, Meerlo P, Havekes R. Recovering object-location memories after sleep deprivation-induced amnesia. Curr Biol. 2023
https://www.cell.com/current-biology/fulltext/S0960-9822(22)01906-6
Quote: “Sleep loss is a common hallmark of modern society, which affects people of all ages1,2 and has a severe impact on body and brain (e.g., Abel et al.,3 Areal et al.,4 Havekes and Aton,5 Holmer et al.,6 Liew and Aung,7 Meerlo et al.,8 and Raven et al.9). Even a brief period of sleep deprivation, following learning and processing of new information, can result in cognitive deficits, particularly in the case of hippocampus-dependent memories.3,10,11,12,13 Several misregulated signaling mechanisms have been identified that may contribute to these cognitive impairments.5,9,10,11,14,15,16 Specifically, previous work by a number of different laboratories demonstrated that sleep deprivation leads to molecular as well as cellular perturbations in the hippocampus.5,17 For instance, a brief period of sleep deprivation decreases the levels of cAMP and PKA, hampers subsequent cAMP/PKA-dependent forms of LTP (e.g., Vecsey et al.18), and leads to structural alterations, resulting in an overall net reduction in the number of dendritic spines in the CA1 and dentate gyrus (DG).19,20,21,22,23 Furthermore, sleep deprivation attenuates mTORC1-dependent protein synthesis in the hippocampus.24 Altogether, these molecular and cellular alterations impacting synaptic plasticity are suggested to be (at least partly) responsible for the cognitive deficits observed after a brief period of sleep deprivation.”
#Brodt S, Inostroza M, Niethard N, Born J. Sleep-A brain-state serving systems memory consolidation. Neuron. 2023
https://www.cell.com/neuron/fulltext/S0896-6273(23)00201-5
Quote: “Although long-term memory consolidation is supported by sleep, it is unclear how it differs from that during wakefulness. Our review, focusing on recent advances in the field, identifies the repeated replay of neuronal firing patterns as a basic mechanism triggering consolidation during sleep and wakefulness. During sleep, memory replay occurs during slow-wave sleep (SWS) in hippocampal assemblies together with ripples, thalamic spindles, neocortical slow oscillations, and noradrenergic activity. Here, hippocampal replay likely favors the transformation of hippocampus-dependent episodic memory into schema-like neocortical memory. REM sleep following SWS might balance local synaptic rescaling accompanying memory transformation with a sleep-dependent homeostatic process of global synaptic renormalization. Sleep-dependent memory transformation is intensified during early development despite the immaturity of the hippocampus. Overall, beyond its greater efficacy, sleep consolidation differs from wake consolidation mainly in that it is supported, rather than impaired, by spontaneous hippocampal replay activity possibly gating memory formation in neocortex.”
#Newbury CR, Crowley R, Rastle K, Tamminen J. Sleep deprivation and memory: Meta-analytic reviews of studies on sleep deprivation before and after learning. Psychol Bull. 2021
https://pmc.ncbi.nlm.nih.gov/articles/PMC8893218/
Quote: “Research suggests that sleep deprivation both before and after encoding has a detrimental effect on memory for newly learned material. However, there is as yet no quantitative analyses of the size of these effects. We conducted two meta-analyses of studies published between 1970 and 2020 that investigated effects of total, acute sleep deprivation on memory (i.e., at least one full night of sleep deprivation): one for deprivation occurring before learning and one for deprivation occurring after learning. The impact of sleep deprivation after learning on memory was associated with Hedges’ g = 0.277, 95% CI [0.177, 0.377]. Whether testing took place immediately after deprivation or after recovery sleep moderated the effect, with significantly larger effects observed in immediate tests. Procedural memory tasks also showed significantly larger effects than declarative memory tasks. The impact of sleep deprivation before learning was associated with Hedges’ g = 0.621, 95% CI [0.473, 0.769]. Egger’s tests for funnel plot asymmetry suggested significant publication bias in both meta-analyses. Statistical power was very low in most of the analyzed studies. Highly powered, preregistered replications are needed to estimate the underlying effect sizes more precisely.”
#Ashton JE, Harrington MO, Langthorne D, Ngo HV, Cairney SA. Sleep deprivation induces fragmented memory loss. Learn Mem. 2020
https://pmc.ncbi.nlm.nih.gov/articles/PMC7079571/
Quote: “Sleep deprivation increases rates of forgetting in episodic memory. Yet, whether an extended lack of sleep alters the qualitative nature of forgetting is unknown. We compared forgetting of episodic memories across intervals of overnight sleep, daytime wakefulness, and overnight sleep deprivation. Item-level forgetting was amplified across daytime wakefulness and overnight sleep deprivation, as compared to sleep. Importantly, however, overnight sleep deprivation led to a further deficit in associative memory that was not observed after daytime wakefulness. These findings suggest that sleep deprivation induces fragmentation among item memories and their associations, altering the qualitative nature of episodic forgetting.”
#Cousins JN, Fernández G. The impact of sleep deprivation on declarative memory. Prog Brain Res. 2019
https://www.sciencedirect.com/science/chapter/bookseries/abs/pii/S007961231930007X?via%3Dihub
Quote: “Sleep plays a crucial role in memory stabilization and integration, yet many people obtain insufficient sleep. This review assesses what is known about the level of sleep deprivation that leads to impairments during encoding, consolidation and retrieval of declarative memories, and what can be determined about the underlying neurophysiological processes. Neuroimaging studies that deprived sleep after learning have provided some of the most compelling evidence for sleep's role in the long-term reorganization of memories in the brain (systems consolidation). However, the behavioral consequences of losing sleep after learning—shown by increased forgetting—appear to recover over time and are unaffected by more common forms of partial sleep restriction across several nights. The capacity to encode new memories is the most vulnerable to sleep loss, since long-term deficits have been observed after total and partial sleep deprivation, while retrieval mechanisms are relatively unaffected. The negative impact of sleep loss on memory has been explored extensively after a night of total sleep deprivation, but further research is needed on the consequences of partial sleep loss over many days so that impairments may be generalized to more common forms of sleep loss.”
#Lo, J.C., Chong, P.L.H., Ganesan, S., Leong, R.L.F. and Chee, M.W.L. Sleep deprivation increases formation of false memory. J Sleep Res (2016)
https://onlinelibrary.wiley.com/doi/10.1111/jsr.12436
Quote: “Retrieving false information can have serious consequences. Sleep is important for memory, but voluntary sleep curtailment is becoming more rampant. Here, the misinformation paradigm was used to investigate false memory formation after 1 night of total sleep deprivation in healthy young adults (N = 58, mean age ± SD = 22.10 ± 1.60 years; 29 males), and 7 nights of partial sleep deprivation (5 h sleep opportunity) in these young adults and healthy adolescents (N = 54, mean age ± SD = 16.67 ± 1.03 years; 25 males). In both age groups, sleep-deprived individuals were more likely than well-rested persons to incorporate misleading post-event information into their responses during memory retrieval (P < 0.050). These findings reiterate the importance of adequate sleep in optimal cognitive functioning, reveal the vulnerability of adolescents' memory during sleep curtailment, and suggest the need to assess eyewitnesses' sleep history after encountering misleading information.”
– To remember you need a cue for the memory, something that is part of the original assembly. A smell, sound, word or maybe the image of an angry crow. Your hippocampus searches its index for the cue, hopefully finding the right stored assembly, and activates it. It fires. Your past experience is retrieved.
Human memory can often recover a complete episode from a partial cue (e.g. beach → whole vacation). This is thought to be driven by a process called “pattern completion” in the hippocampus.
#Miehl, C., Onasch, S., Festa, D. and Gjorgjieva, J. Formation and computational implications of assemblies in neural circuits. J Physiol. (2023)
https://doi.org/10.1113/JP282750
Quote: “The idea of assemblies, or ensembles, as the basic units of cognition has recently replaced the neuron-centric view (Buzsáki, 2010; Eichenbaum, 2018; Huyck & Passmore, 2013; Yuste, 2015). An emergent core assumption from this framework is that each ensemble represents a specific concept or feature, acting as the fundamental unit for memory storage (Neves et al., 2008). One advantage of having strong recurrent connectivity (assembly) compared to only considering correlated rates (ensemble) is stimulus amplification (Peron et al., 2020), thus enabling weaker stimuli to elicit a recognisable response and to increase robustness whereby the malfunction or death of single neurons or synapses will not affect the represented concept. Consequently, an incomplete stimulus can be sufficient to evoke the complete assembly – a phenomenon named pattern completion, which is especially relevant for memory retrieval (Guzman et al., 2016). Moreover, recurrent interactions alone may be sufficient to keep an assembly active after stimulation, thus enabling the brain to decouple intrinsic activity from external stimulation and modify learned concepts independently from specific external inputs (Harris, 2005).”
#Grande X, Berron D, Horner AJ, Bisby JA, Düzel E, Burgess N. Holistic Recollection via Pattern Completion Involves Hippocampal Subfield CA3. J Neurosci. 2019
https://pmc.ncbi.nlm.nih.gov/articles/PMC6786823/
Quote: “Memories of daily events usually involve multiple elements, although a single element can be sufficient to prompt recollection of the whole event. Such holistic recollection is thought to require reactivation of brain activity representing the full event from one event element (“pattern completion”). Computational and animal models suggest that mnemonic pattern completion is accomplished in a specific subregion of the hippocampus called CA3, but empirical evidence in humans was lacking. Here, we leverage the ultra-high resolution of 7 tesla neuroimaging to provide first evidence for a strong involvement of the human CA3 in holistic recollection of multi-element events via pattern completion.”
#Staresina BP, Wimber M. A Neural Chronometry of Memory Recall. Trends Cogn Sci. 2019
https://www.cell.com/trends/cognitive-sciences/fulltext/S1364-6613(19)30235-9
Quote: “Episodic memory allows us to mentally travel through time. How does the brain convert a simple reminder cue into a full-blown memory of past events and experiences? In this review, we integrate recent developments in the cognitive neuroscience of human memory retrieval, pinpointing the neural chronometry underlying successful recall. Electrophysiological recordings suggest that sensory cues proceed into the medial temporal lobe within the first 500 ms. At this point, a hippocampal process sets in, geared toward internal pattern completion and coordination of cortical memory reinstatement between 500 and 1500 ms. We further highlight the dynamic principles governing the recall process, which include a reversal of perceptual information flows, temporal compression, and theta clocking.”
– Which means that as you experience the memory in your mind, the neurons involved are bathed in chemicals that make them able to change their structure again.
Retrieving/recalling a consolidated (i.e., stored) memory can return it to a labile state that requires reconsolidation, during which it can be strengthened, weakened, or modified. In episodic memory, partial reminders or old-context/new-content pairings can drive the incorporation of new information into existing memories. This “updating” supports flexible world models, but the same processes underlie distortions and false memories.
It is important to note that some memories, for example very strong and/or emotional ones, appear to be more resistant to this kind of change, but the reasons are not yet well understood.
Neuromodulators such as e.g. dopamine are thought to play a major role in reconsolidation and memory updating.
#O'Neill OS, Winters BD. Breaking boundaries: Dopamine's role in prediction error, salient novelty, and memory reconsolidation. Neuroscience. 2026
https://www.sciencedirect.com/science/article/pii/S0306452225011972
Quote: “For memories to remain relevant and adaptive over the lifespan, modifications under specific conditions are required. Memory reconsolidation theory suggests that when a memory is reactivated, it can become labile, a state known as destabilization. This process is regulated by complex and dynamic neurobiological changes representing biological boundary conditions, which likely protect important memories from undergoing unnecessary or potentially maladaptive modifications. External cues, such as prediction error or other forms of salient novel information, can promote destabilization of these resistant memory traces. Accordingly, various neurobiological mechanisms related to the signaling of prediction errors and salient novelty have been implicated in overcoming boundary conditions, permitting memory modification. Here, we review the existing literature regarding the mechanisms for overcoming biological boundary conditions, with specific focus on the role of the neurotransmitter dopamine and its well documented functions related to prediction error, novelty detection, and memory reconsolidation. We aim to describe the nuanced role of dopamine in these processes as it pertains to destabilizing modification-resistant memories, highlight potential interactions with alternate neurotransmitter systems for this process, and bridge findings from reward learning and novelty processing to convey a holistic view of dopamine’s role in memory reconsolidation more broadly.”
#Bellfy L, Kwapis JL. Molecular Mechanisms of Reconsolidation-Dependent Memory Updating. International Journal of Molecular Sciences. 2020
https://www.mdpi.com/1422-0067/21/18/6580
Quote: “Memory is not a stable record of experience, but instead is an ongoing process that allows existing memories to be modified with new information through a reconsolidation-dependent updating process. For a previously stable memory to be updated, the memory must first become labile through a process called destabilization. Destabilization is a protein degradation-dependent process that occurs when new information is presented. Following destabilization, a memory becomes stable again through a protein synthesis-dependent process called restabilization. Much work remains to fully characterize the mechanisms that underlie both destabilization and subsequent restabilization, however. In this article, we briefly review the discovery of reconsolidation as a potential mechanism for memory updating. We then discuss the behavioral paradigms that have been used to identify the molecular mechanisms of reconsolidation-dependent memory updating. Finally, we outline what is known about the molecular mechanisms that support the memory updating process. Understanding the molecular mechanisms underlying reconsolidation-dependent memory updating is an important step toward leveraging this process in a therapeutic setting to modify maladaptive memories and to improve memory when it fails.”
#Osorio-Gómez D, Miranda MI, Guzmán-Ramos K and Bermúdez-Rattoni F (2023) Transforming experiences: Neurobiology of memory updating/editing. Front. Syst. Neurosci.
https://www.frontiersin.org/journals/systems-neuroscience/articles/10.3389/fnsys.2023.1103770/full
Quote: “Long-term memory is achieved through a consolidation process where structural and molecular changes integrate information into a stable memory. However, environmental conditions constantly change, and organisms must adapt their behavior by updating their memories, providing dynamic flexibility for adaptive responses. Consequently, novel stimulation/experiences can be integrated during memory retrieval; where consolidated memories are updated by a dynamic process after the appearance of a prediction error or by the exposure to new information, generating edited memories. This review will discuss the neurobiological systems involved in memory updating including recognition memory and emotional memories. In this regard, we will review the salient and emotional experiences that promote the gradual shifting from displeasure to pleasure (or vice versa), leading to hedonic or aversive responses, throughout memory updating. Finally, we will discuss evidence regarding memory updating and its potential clinical implication in drug addiction, phobias, and post-traumatic stress disorder.”
#Haubrich, J., Nader, K. (2016). Memory Reconsolidation. In: Clark, R.E., Martin, S. (eds) Behavioral Neuroscience of Learning and Memory. Current Topics in Behavioral Neurosciences, vol 37. Springer, Cham.
https://doi.org/10.1007/7854_2016_463
Quote: “Scientific advances in the last decades uncovered that memory is not a stable, fixed entity. Apparently stable memories may become transiently labile and susceptible to modifications when retrieved due to the process of reconsolidation. Here, we review the initial evidence and the logic on which reconsolidation theory is based, the wide range of conditions in which it has been reported and recent findings further revealing the fascinating nature of this process. Special focus is given to conceptual issues of when and why reconsolidation happen and its possible outcomes. Last, we discuss the potential clinical implications of memory modifications by reconsolidation.”
#Sinclair, Alyssa H. et al. Prediction Error and Memory Reactivation: How Incomplete Reminders Drive Reconsolidation. Trends in Neurosciences (2019)
https://www.cell.com/trends/neurosciences/abstract/S0166-2236(19)30151-1
Quote: “Memories are readily distorted. What conditions allow memories to be altered? Converging evidence implicates prediction error, or surprise, as a key mechanism that renders memories malleable. Recent reconsolidation studies have used incomplete reminders to elicit prediction error; retrieval cues that partially replicate an encoding experience allow memories to be distorted, updated, and strengthened. Here, we review diverse evidence that incomplete reminders govern human memory updating, ranging from classical conditioning to naturalistic episodes. Through the unifying theme of predictive coding, we discuss evidence from reconsolidation theory and nonmonotonic plasticity. We argue that both animal and human reconsolidation research can benefit from critically examining prediction error and incomplete reminders. These findings bear implications for pathological fear memories, false memories, misinformation, and education.”
#Jardine KH, Huff AE, Wideman CE, McGraw SD, Winters BD. The evidence for and against reactivation-induced memory updating in humans and nonhuman animals. Neurosci Biobehav Rev. 2022
https://www.sciencedirect.com/science/article/abs/pii/S0149763422000872?via%3Dihub
Quote: “Systematic investigation of reactivation-induced memory updating began in the 1960s, and a wave of research in this area followed the seminal articulation of “reconsolidation” theory in the early 2000s. Myriad studies indicate that memory reactivation can cause previously consolidated memories to become labile and sensitive to weakening, strengthening, or other forms of modification. However, from its nascent period to the present, the field has been beset by inconsistencies in researchers’ abilities to replicate seemingly established effects. Here we review these many studies, synthesizing the human and nonhuman animal literature, and suggest that these failures-to-replicate reflect a highly complex and delicately balanced memory modification system, the substrates of which must be finely tuned to enable adaptive memory updating while limiting maladaptive, inaccurate modifications. A systematic approach to the entire body of evidence, integrating positive and null findings, will yield a comprehensive understanding of the complex and dynamic nature of long-term memory storage and the potential for harnessing modification processes to treat mental disorders driven by pervasive maladaptive memories.”
#Wideman CE, Jardine KH, Winters BD. Involvement of classical neurotransmitter systems in memory reconsolidation: Focus on destabilization. Neurobiol Learn Mem. 2018
https://www.sciencedirect.com/science/article/abs/pii/S1074742718302569?via%3Dihub
Quote: “When consolidated long-term memories are reactivated they can destabilize, rendering the memory labile and vulnerable to modification. This period of lability is followed by reconsolidation, a process that restabilizes the memory trace. Reactivation-induced memory destabilization is the gateway process to reconsolidation, but research in this area has focused primarily on the mechanisms underlying post-reactivation restabilization. As a result, our understanding of processes subserving destabilization have lagged behind those responsible for reconsolidation. Here we review the literature investigating the neural basis of reactivation-induced memory destabilization. We begin by reviewing memory destabilization broadly and the boundary conditions that influence the likelihood of reactivated memories to destabilize. We then discuss the fact that boundary conditions can be overcome in the presence of novelty, providing evidence for the theory that reconsolidation is a mechanism for memory updating. From here, we delve into a detailed review of the role of classical neurotransmitter systems, including dopamine, serotonin, noradrenaline, glutamate, GABA and acetylcholine, in reconsolidation, with a focus on their involvement in destabilization. Many of these neurotransmitters appear capable of promoting memory destabilization, and research investigating the cellular pathways through which they influence destabilization is a growing area. However, gaps remain in our understanding of how these neurotransmitters work in conjunction with one another to support destabilization across different types of memory and in different brain regions. Advances in the coming years within this research field should greatly contribute to our understanding of the neural mechanisms that influence the dynamic process of long-term memory storage and modification, information crucial to the development of potential treatments for disorders characterized by strong, maladaptive memories.”
– Your hippocampus organizes memories based on a few things but most importantly: the context of the experience. And the context is now very different. When you formed the memory you were tired, in a mildly bad mood before work and very surprised. Right now you are pulling the memory as you are having a night out with your friends and are telling them your hilarious story of the epic squirrel crow fight. And this new context seeps into the memory.
If a memory is recalled in a new context, e.g. in a different emotional state, with new information, or an altered interpretation, that new information can be incorporated into the memory trace in the brain, partially rewriting it. This process is researched particularly in the context of emotional learning and fear conditioning, psychotherapy (e.g. exposure therapy, trauma treatment), and habit learning.
It is important to note that some memories, for example very strong and/or emotional ones, appear to be more resistant to this kind of change, but the reasons for this are not yet well understood.
#Bayer H, Bertoglio LJ, Maren S, Stern CAJ. Windows of change: Revisiting temporal and molecular dynamics of memory reconsolidation and persistence. Neurosci Biobehav Rev. 2025
https://www.sciencedirect.com/science/article/abs/pii/S0149763425001988
Quote: “Numerous studies have shown that, under specific conditions, memory retrieval and reactivation place memory into a labile state, initiating a new re-stabilization phase known as reconsolidation (Przybyslawski and Sara, 1997, Nader et al., 2000; Pedreira et al., 2004; Bustos et al., 2009; Stern et al., 2012; Franzen et al., 2019; Merlo et al., 2014). [...] Memory reconsolidation has been demonstrated in several species, including mice, rats, zebrafish, pond snails, crabs, sea slugs, non-human primates, and humans (Lee et al., 2017). This underscores the evolutionary conservation of this process, highlighting its fundamental role in maintaining the relevance of memory content in a constantly changing environment.
Once a memory is destabilized, various cellular processes must be recruited to restabilize the memory for persistent storage (Tronson and Taylor, 2007, Jarome and Lubin, 2014; Kida, 2019). During this period of instability, however, the memory can be updated, strengthened, weakened, or even disrupted through various interventions, such as pharmacological agents and behavioral procedures. The susceptibility of long-term memory to manipulation has sparked interest in targeting memory reconsolidation therapeutically. There has been considerable effort to explore whether reconsolidation update procedures might be used as interventions for PTSD or substance use disorders, for example. Additionally, understanding whether reconsolidation processes contribute to memory impairments in Alzheimer's disease might also lead to the development of novel interventions to slow cognitive decline (Cain et al., 2012, Schwabe et al., 2014, Kroes et al., 2016, Beckers and Kindt, 2017, Lee et al., 2017, Haubrich and Nader, 2018, Walsh et al., 2018, Deng et al., 2023, Huang et al., 2024, Lattal, 2024, Merlo et al., 2024).
Despite considerable advances in understanding memory reconsolidation, fundamental questions remain regarding the nature, duration, and neurobiological mechanisms that stabilize long-term memory after retrieval and reactivation.”
Speer, M.E., Ibrahim, S., Schiller, D. et al. Finding positive meaning in memories of negative events adaptively updates memory. Nat Commun (2021). https://doi.org/10.1038/s41467-021-26906-4
Quote: “Finding positive meaning in past negative memories is associated with enhanced mental health. Yet it remains unclear whether it leads to updates in the memory representation itself. Since memory can be labile after retrieval, this leaves the potential for modification whenever its reactivated. Across four experiments, we show that positively reinterpreting negative memories adaptively updates them, leading to the re-emergence of positivity at future retrieval. Focusing on the positive aspects after negative recall leads to enhanced positive emotion and changes in memory content during recollection one week later, remaining even after two months. Consistent with a reactivation-induced reconsolidation account, memory updating occurs only after a reminder and twenty four hours, but not a one hour delay. Multi-session fMRI showed adaptive updates are reflected in greater hippocampal and ventral striatal pattern dissimilarity across retrievals. This research highlights the mechanisms by which updating of maladaptive memories occurs through a positive emotion-focused strategy.”
#D. Stawarczyk, C.N. Wahlheim, J.A. Etzel, A.Z. Snyder, & J.M. Zacks, Aging and the encoding of changes in events: The role of neural activity pattern reinstatement, Proc. Natl. Acad. Sci. U.S.A. (2020)
https://doi.org/10.1073/pnas.1918063117
Quote: “Memory updating upon change detection has been found to depend crucially on interactions between the hippocampus, the surrounding medial temporal lobes (MTL), and the rest of the cortex (6, 7). Memory updating comprises several computational operations with different neural correlates and behavioral signatures (8, 9). These include pattern completion, which is the prediction function that activates relevant prior memories and knowledge based on environmental cues; pattern separation and differentiation, which keep features of the two experiences separate; and integration, which captures the relationships between different features of similar events. In order to integrate memory representations of events that are similar but include discrepant features, the brain needs to register the discrepancy and use it to prompt new learning. Models of memory updating propose that, when things change, pattern completion leads to prediction errors that can drive new learning, including integration processes to form configural memory representations (5, 10).
[...]
Neuroimaging data indicate that patterns of brain activity present while encoding new information are reinstated when this information is recollected, both for simple laboratory materials (e.g., refs. 16 and 17) and for more complex stimuli such as movies of everyday activities (18, 19). This effect is usually the strongest in the posterior areas of the default network (DN) (20), more specifically part of the posteromedial cortex (PMC) that includes the posterior cingulate cortex (PCC) and retrosplenial cortex (Rsp), and in the medial temporal lobe (MTL), including the parahippocampal cortex (PHC) and hippocampus. These regions are sometimes referred to as the posterior medial system (21) or contextual association network (22) due to their strong involvement in long-term memory recollection, particularly when episodic representations of everyday events must be remembered from visual cues (13, 23).”
#O'Neill OS, Winters BD. Breaking boundaries: Dopamine's role in prediction error, salient novelty, and memory reconsolidation. Neuroscience. 2026
https://www.sciencedirect.com/science/article/pii/S0306452225011972
Quote: “For memories to remain relevant and adaptive over the lifespan, modifications under specific conditions are required. Memory reconsolidation theory suggests that when a memory is reactivated, it can become labile, a state known as destabilization. This process is regulated by complex and dynamic neurobiological changes representing biological boundary conditions, which likely protect important memories from undergoing unnecessary or potentially maladaptive modifications. External cues, such as prediction error or other forms of salient novel information, can promote destabilization of these resistant memory traces. Accordingly, various neurobiological mechanisms related to the signaling of prediction errors and salient novelty have been implicated in overcoming boundary conditions, permitting memory modification. Here, we review the existing literature regarding the mechanisms for overcoming biological boundary conditions, with specific focus on the role of the neurotransmitter dopamine and its well documented functions related to prediction error, novelty detection, and memory reconsolidation. We aim to describe the nuanced role of dopamine in these processes as it pertains to destabilizing modification-resistant memories, highlight potential interactions with alternate neurotransmitter systems for this process, and bridge findings from reward learning and novelty processing to convey a holistic view of dopamine’s role in memory reconsolidation more broadly.”
#Lee JLC, Nader K, Schiller D. An Update on Memory Reconsolidation Updating. Trends Cogn Sci. 2017
https://www.cell.com/trends/cognitive-sciences/abstract/S1364-6613(17)30078-5
Quote: “The reactivation of a stored memory in the brain can make the memory transiently labile. During the time it takes for the memory to restabilize (reconsolidate) the memory can either be reduced by an amnesic agent or enhanced by memory enhancers. The change in memory expression is related to changes in the brain correlates of long-term memory. Many have suggested that such retrieval-induced plasticity is ideally placed to enable memories to be updated with new information. This hypothesis has been tested experimentally, with a translational perspective, by attempts to update maladaptive memories to reduce their problematic impact. We review here progress on reconsolidation updating studies, highlighting their translational exploitation and addressing recent challenges to the reconsolidation field.”
#Meir Drexler S, Wolf OT. Behavioral disruption of memory reconsolidation: From bench to bedside and back again. Behav Neurosci. 2018
Quote: “During the postretrieval reconsolidation “window”, memories can be disrupted, strengthened, or updated using various pharmacological and behavioral manipulations. Behavioral manipulations are more ecologically valid, thus allowing better understating of memory modification under natural conditions, but they can also be less potent compared to pharmacological interventions. In this review we present the current human and animal literature, aiming to understand the modulatory factors (i.e., task relevance, complexity, intensity) that promote reconsolidation disruption in purely behavioral means. The reviewed studies have suggested that both very simple tasks and more complex learning paradigms can be used to disrupt or update memory reconsolidation, even of stronger emotional memories. Stress exposure is a possible interference task, yet the conflicting results leave many open questions regarding its required timing and intensity. Going from bench to bedside and back again, we point to the need for more research in clinical populations to establish the therapeutic potential of reconsolidation-based treatments. Several findings from outside the laboratory offer promising leads for future research.”
#Monfils MH, Holmes EA. Memory boundaries: opening a window inspired by reconsolidation to treat anxiety, trauma-related, and addiction disorders. Lancet Psychiatry. 2018
https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(18)30270-0/abstract
Quote: “Pioneering research over the past two decades has shown that memories are far more malleable than we once thought, thereby highlighting the potential for new clinical avenues for treatment of psychopathology. We first briefly review the historical foundation of memory reconsolidation—a concept that refers to hypothetical processes that occur when a memory is retrieved and restored. Then, we provide an overview of the basic research on memory reconsolidation that has been done with humans and other animals, focusing on models of fear, anxiety-related disorders, and addiction, from the perspective that they all involve disorders of memory. This basic research has fuelled early stage developments of novel treatment techniques. More specifically, we consider behavioural interventions inspired by reconsolidation updating, namely retrieval-extinction techniques. We discuss the set of principles that would be needed for memory modifications within a putative reconsolidation time window, and review research that employs reconsolidation-based strategies with clinical populations. We conclude by highlighting current pitfalls and controversies surrounding the use of reconsolidation-based approaches, but end on an optimistic note for clinical research going forward. Despite the challenges, we believe that drawing on ideas from psychological science can help open up treatment innovation.”
#Paul T, Asthana MK. Investigating the effect of memory strength within reactivation-extinction paradigm on cue-dependent amnesia. Neuroscience. 2025
https://www.ibroneuroscience.org/article/S0306-4522(25)00777-8/abstract
Quote: “Strong emotional memories serve as core symptoms of major psychological disorders. Paradigms based on memory reconsolidation have demonstrated superior results compared to standard methods. However, many studies have failed to replicate the findings in humans, highlighting the potential boundary conditions of memory, such as the strength of memory and prediction error (PE) that limits the effectiveness of the retrieval extinction paradigm. In the current study, we use a three-day fear conditioning paradigm. Three geometrical figures (CSs) were paired with two aversive tones and images (US) at partial (50%) and continuous (100%) reinforcement rates. The results showed that SCR generated higher conditioned responses during acquisition for continuous reinforcement than partial reinforcement. We find that there was a general decrease in SCR of fear reinstatement in groups treated with PE than standard methods. Also, group-wise comparison shows that fear returns in 100% reinforcement even if treated with PE in comparison to 50% reinforcement. The results demonstrate that memory strength and PE serve as boundary conditions for memory reconsolidation. This data is significant as it is the first study to examine memory strength within the so-called reactivation-extinction procedure in the context of the reconsolidation framework in human subjects.”
#Zhang JJ, Haubrich J, Bernabo M, Finnie PSB, Nader K. Limits on lability: Boundaries of reconsolidation and the relationship to metaplasticity. Neurobiol Learn Mem. 2018
https://www.sciencedirect.com/science/article/abs/pii/S1074742718300406?via%3Dihub
Quote: “Reconsolidation, a process by which long-term memories are rendered malleable following retrieval, has been shown to occur across many different species and types of memory. However, there are conditions under which memories do not reconsolidate, and the reasons for this are poorly understood. One emerging theory is that these boundary conditions are mediated by a form of metaplasticity: cellular changes through which experience can affect future synaptic plasticity. We review evidence that N-methyl-D-aspartate receptors (NMDARs) might contribute to this phenomenon, and hypothesize that resistance to memory destabilization may be mediated by the ratio of GluN2A/GluN2B subunits that make up these receptors. Qualities such as memory strength and the age of the memory may increase the GluN2A/GluN2B ratio, reducing the ability of reactivation cues to induce destabilization, thereby preventing reconsolidation. Other examples of experience-dependent learning and evolutionary perspectives of reconsolidation are also discussed.”
– New connections form, some synapses are weakened while others reconfigure.
Here we are referencing synaptic plasticity, i.e. the brain's ability to change and adapt by modifying the strength of connections (synapses) between neurons. When stored memories are recalled, they can become temporarily unstable and must undergo a process called “reconsolidation” to persist. During reconsolidation, small changes may get introduced to the memory trace in the brain. This lability is linked to synaptic plasticity mechanisms that can both weaken and restabilize synapses.
While reconsolidation is also described and researched in humans, most of these mechanistic details are inferred from findings in mammals, e.g. rats.
#Pauli Q, Bonin RP. Making memories malleable: The role and regulation of synaptic depotentiation in engram alteration. Neurosci Biobehav Rev. 2026
https://www.sciencedirect.com/science/article/pii/S0149763425004828?via%3Dihub
Quote: “Activity-dependent increases in synaptic strength, including long-term potentiation, are involved in learning to facilitate memory encoding and recall. The reversal of long-term potentiation, termed synaptic depotentiation, has also been observed in brain slices and in behaving animals; yet its precise role in memory processing is unclear. Here, we review functional and structural evidence for synaptic depotentiation in memory processes that support flexible and adaptive behaviour in rodents. We present evidence that depotentiation serves to weaken synapses potentiated during prior learning events to facilitate active forgetting and memory destabilization. The demonstrated ability for prior synaptic activity and neuromodulatory inputs to regulate depotentiation may contribute to scenarios where certain memories resist forgetting or modification. Understanding the synaptic mechanisms that give rise to memory flexibility enables the development of better in vitro models to provide a more accurate conceptualization of the nature of forgetting and indicate potential new treatment avenues for memory-related disorders.”
#Bonin RP, De Koninck Y. Reconsolidation and the regulation of plasticity: moving beyond memory. Trends Neurosci. 2015
https://www.cell.com/trends/neurosciences/abstract/S0166-2236(15)00097-1
Quote: “Memory reconsolidation is a protein synthesis-dependent process that preserves, in some form, memories that have been destabilized through recall. Reconsolidation is a nearly universal phenomenon, occurring in a diverse array of species and learning tasks. The function of reconsolidation remains unclear but it has been proposed as a mechanism for updating or strengthening memories. Observations of an analog of reconsolidation in vitro and in sensory systems indicate that reconsolidation is unlikely to be a learning-specific phenomenon and may serve a broader function. We propose that reconsolidation arises from the activity-dependent induction of two coincident but opposing processes: the depotentiation and repotentiation of strengthened synapses. These processes suggest that reconsolidation reflects a fundamental mechanism that regulates and preserves synaptic strength.”
– And it is the same with all your memories – when you retrieve them, your brain adds new information or forgets some, and incorporates your emotions and expectations. In a sense it updates your past life to fit the narrative of your present life. Over time, even core memories can shift, combine with others, or generate entirely new elements.
If a memory is recalled in a new context, e.g. in a different emotional state, new information, or altered interpretation, that new information can be incorporated into the memory trace in the brain, partially rewriting it. This process is researched particularly in the context of emotional learning and fear conditioning, psychotherapy (e.g. exposure therapy, trauma treatment), and habit learning.
It is important to note that some memories, for example very strong and/or emotional ones, appear to be more resistant to this kind of change, but the reasons are not yet well understood.
#Bayer H, Bertoglio LJ, Maren S, Stern CAJ. Windows of change: Revisiting temporal and molecular dynamics of memory reconsolidation and persistence. Neurosci Biobehav Rev. 2025
https://www.sciencedirect.com/science/article/abs/pii/S0149763425001988
Quote: “Numerous studies have shown that, under specific conditions, memory retrieval and reactivation place memory into a labile state, initiating a new re-stabilization phase known as reconsolidation (Przybyslawski and Sara, 1997, Nader et al., 2000; Pedreira et al., 2004; Bustos et al., 2009; Stern et al., 2012; Franzen et al., 2019; Merlo et al., 2014). [...] Memory reconsolidation has been demonstrated in several species, including mice, rats, zebrafish, pond snails, crabs, sea slugs, non-human primates, and humans (Lee et al., 2017). This underscores the evolutionary conservation of this process, highlighting its fundamental role in maintaining the relevance of memory content in a constantly changing environment.
Once a memory is destabilized, various cellular processes must be recruited to restabilize the memory for persistent storage (Tronson and Taylor, 2007, Jarome and Lubin, 2014; Kida, 2019). During this period of instability, however, the memory can be updated, strengthened, weakened, or even disrupted through various interventions, such as pharmacological agents and behavioral procedures. The susceptibility of long-term memory to manipulation has sparked interest in targeting memory reconsolidation therapeutically. There has been considerable effort to explore whether reconsolidation update procedures might be used as interventions for PTSD or substance use disorders, for example. Additionally, understanding whether reconsolidation processes contribute to memory impairments in Alzheimer's disease might also lead to the development of novel interventions to slow cognitive decline (Cain et al., 2012, Schwabe et al., 2014, Kroes et al., 2016, Beckers and Kindt, 2017, Lee et al., 2017, Haubrich and Nader, 2018, Walsh et al., 2018, Deng et al., 2023, Huang et al., 2024, Lattal, 2024, Merlo et al., 2024).
Despite considerable advances in understanding memory reconsolidation, fundamental questions remain regarding the nature, duration, and neurobiological mechanisms that stabilize long-term memory after retrieval and reactivation.”
#Lee JLC, Nader K, Schiller D. An Update on Memory Reconsolidation Updating. Trends Cogn Sci. 2017
https://www.cell.com/trends/cognitive-sciences/abstract/S1364-6613(17)30078-5
Quote: “The reactivation of a stored memory in the brain can make the memory transiently labile. During the time it takes for the memory to restabilize (reconsolidate) the memory can either be reduced by an amnesic agent or enhanced by memory enhancers. The change in memory expression is related to changes in the brain correlates of long-term memory. Many have suggested that such retrieval-induced plasticity is ideally placed to enable memories to be updated with new information. This hypothesis has been tested experimentally, with a translational perspective, by attempts to update maladaptive memories to reduce their problematic impact. We review here progress on reconsolidation updating studies, highlighting their translational exploitation and addressing recent challenges to the reconsolidation field.”
#Meir Drexler S, Wolf OT. Behavioral disruption of memory reconsolidation: From bench to bedside and back again. Behav Neurosci. 2018
Quote: “During the postretrieval reconsolidation “window”, memories can be disrupted, strengthened, or updated using various pharmacological and behavioral manipulations. Behavioral manipulations are more ecologically valid, thus allowing better understating of memory modification under natural conditions, but they can also be less potent compared to pharmacological interventions. In this review we present the current human and animal literature, aiming to understand the modulatory factors (i.e., task relevance, complexity, intensity) that promote reconsolidation disruption in purely behavioral means. The reviewed studies have suggested that both very simple tasks and more complex learning paradigms can be used to disrupt or update memory reconsolidation, even of stronger emotional memories. Stress exposure is a possible interference task, yet the conflicting results leave many open questions regarding its required timing and intensity. Going from bench to bedside and back again, we point to the need for more research in clinical populations to establish the therapeutic potential of reconsolidation-based treatments. Several findings from outside the laboratory offer promising leads for future research.”
#Monfils MH, Holmes EA. Memory boundaries: opening a window inspired by reconsolidation to treat anxiety, trauma-related, and addiction disorders. Lancet Psychiatry. 2018
https://www.thelancet.com/journals/lanpsy/article/PIIS2215-0366(18)30270-0/abstract
Quote: “Pioneering research over the past two decades has shown that memories are far more malleable than we once thought, thereby highlighting the potential for new clinical avenues for treatment of psychopathology. We first briefly review the historical foundation of memory reconsolidation—a concept that refers to hypothetical processes that occur when a memory is retrieved and restored. Then, we provide an overview of the basic research on memory reconsolidation that has been done with humans and other animals, focusing on models of fear, anxiety-related disorders, and addiction, from the perspective that they all involve disorders of memory. This basic research has fuelled early stage developments of novel treatment techniques. More specifically, we consider behavioural interventions inspired by reconsolidation updating, namely retrieval-extinction techniques. We discuss the set of principles that would be needed for memory modifications within a putative reconsolidation time window, and review research that employs reconsolidation-based strategies with clinical populations. We conclude by highlighting current pitfalls and controversies surrounding the use of reconsolidation-based approaches, but end on an optimistic note for clinical research going forward. Despite the challenges, we believe that drawing on ideas from psychological science can help open up treatment innovation.”
#Paul T, Asthana MK. Investigating the effect of memory strength within reactivation-extinction paradigm on cue-dependent amnesia. Neuroscience. 2025
https://www.ibroneuroscience.org/article/S0306-4522(25)00777-8/abstract
Quote: “Strong emotional memories serve as core symptoms of major psychological disorders. Paradigms based on memory reconsolidation have demonstrated superior results compared to standard methods. However, many studies have failed to replicate the findings in humans, highlighting the potential boundary conditions of memory, such as the strength of memory and prediction error (PE) that limits the effectiveness of the retrieval extinction paradigm. In the current study, we use a three-day fear conditioning paradigm. Three geometrical figures (CSs) were paired with two aversive tones and images (US) at partial (50%) and continuous (100%) reinforcement rates. The results showed that SCR generated higher conditioned responses during acquisition for continuous reinforcement than partial reinforcement. We find that there was a general decrease in SCR of fear reinstatement in groups treated with PE than standard methods. Also, group-wise comparison shows that fear returns in 100% reinforcement even if treated with PE in comparison to 50% reinforcement. The results demonstrate that memory strength and PE serve as boundary conditions for memory reconsolidation. This data is significant as it is the first study to examine memory strength within the so-called reactivation-extinction procedure in the context of the reconsolidation framework in human subjects.”
– Your memory system is deeply intertwined with the mechanisms for learning – it was never designed to create an accurate representation of the world.
Learning and memory are not separate modules or functions in your brain, but different phases of a single adaptive process. We are simplifying a lot here, but in a nutshell: learning changes the brain, and memory is the persistence and flexible reuse of those changes.
From an expert on neuroscience we have also received the following insight: “[Y]our memory is not just for you to remember a specific event (e.g., the answer to one specific math problem), but for you to remember generalizable categories and strategies (e.g., the method to solve a type of math problem). So in a way it's good our memories aren't overly specific, so that we can apply them to similar but new situations that we encounter."
#Battaglia S, Avenanti A, Vécsei L, Tanaka M. Neural Correlates and Molecular Mechanisms of Memory and Learning. International Journal of Molecular Sciences. 2024
https://doi.org/10.3390/ijms25052724
Quote: “Memory and learning are essential cognitive processes that enable us to obtain, retain, and recall information. These factors are crucial for survival, adaptation, and creativity. However, the neural and molecular mechanisms that underlie these cognitive functions are not fully elucidated. For decades, researchers have been fascinated by the neurobiological and molecular basis of acquiring, storing, and retrieving information [1]. Recent neuroimaging technologies have provided valuable insights into underlying neuroanatomical brain circuits [2,3,4,5,6,7]. The amygdala, hippocampus, and prefrontal cortex (PFC) are pivotal for shaping memory and facilitating learning. The amygdala, recognized for its significance in emotional processing, interacts with downstream structures such as the hypothalamus and brainstem regions, influencing the expression of emotionally charged responses [8,9,10]. The inhibitory mechanisms within the amygdala, including specific divisions and nuclei, contribute to memory modulation. The hippocampus, which is essential for spatial navigation and contextual memory, forms direct projections with the infralimbic cortex in the PFC and the basolateral amygdala [11,12]. Distinct subregions of the hippocampus have been implicated in various human behavioral features, highlighting their multifaceted roles in cognitive processes.”
#Ortega-de San Luis C, Ryan TJ. Understanding the physical basis of memory: Molecular mechanisms of the engram. J Biol Chem. 2022
https://pmc.ncbi.nlm.nih.gov/articles/PMC9065729/
Quote: “Memory, defined as the storage and use of learned information in the brain, is necessary to modulate behavior and critical for animals to adapt to their environments and survive. Despite being a cornerstone of brain function, questions surrounding the molecular and cellular mechanisms of how information is encoded, stored, and recalled remain largely unanswered. One widely held theory is that an engram is formed by a group of neurons that are active during learning, which undergoes biochemical and physical changes to store information in a stable state, and that are later reactivated during recall of the memory. In the past decade, the development of engram labeling methodologies has proven useful to investigate the biology of memory at the molecular and cellular levels. Engram technology allows the study of individual memories associated with particular experiences and their evolution over time, with enough experimental resolution to discriminate between different memory processes: learning (encoding), consolidation (the passage from short-term to long-term memories), and storage (the maintenance of memory in the brain). Here, we review the current understanding of memory formation at a molecular and cellular level by focusing on insights provided using engram technology.
[...]
In 1904, Richard Semon (7) proposed the idea of the “engram” and defined it as “the enduring though primary latent modification in the irritable substance produced by a stimulus (from an experience).” An engram, sometimes understood as a synonym for memory trace, is formed by a group of neurons that (1) become activated by a specific learning experience, (2) are modified by this experience, and (3) are reactivated by re-exposure to the same experience, inducing a change in the behavior of the animal (8). Engram cells, therefore, are at least a part of the physical place or substrate where learning leaves imprints in the brain. Sets of engram cells can be found sparse in many areas of the brain, forming an engrome, or engram complex (9).
[...]
Learning is the process of acquiring new information that culminates in the creation of a memory.”
– So your memories are updated as you gain new experiences and information. Ironically, the more you remember something actively, the less of the original experience remains.
Repeated recalling of a memory does more than just “pull out” a stored record; it actively rewrites the memory. The question of whether a memory is still “the original memory” after dozens (or hundreds) of times of being rewritten into your brain is partially also a philosophical problem. Meanwhile, the relationship between repeated recalls and accuracy is highly context-dependent: research shows that repeated retrieval can both stabilize and distort memories.
#Osorio-Gómez D, Miranda MI, Guzmán-Ramos K and Bermúdez-Rattoni F (2023) Transforming experiences: Neurobiology of memory updating/editing. Front. Syst. Neurosci.
https://www.frontiersin.org/journals/systems-neuroscience/articles/10.3389/fnsys.2023.1103770/full
Quote: “Long-term memory is achieved through a consolidation process where structural and molecular changes integrate information into a stable memory. However, environmental conditions constantly change, and organisms must adapt their behavior by updating their memories, providing dynamic flexibility for adaptive responses. Consequently, novel stimulation/experiences can be integrated during memory retrieval; where consolidated memories are updated by a dynamic process after the appearance of a prediction error or by the exposure to new information, generating edited memories. This review will discuss the neurobiological systems involved in memory updating including recognition memory and emotional memories. In this regard, we will review the salient and emotional experiences that promote the gradual shifting from displeasure to pleasure (or vice versa), leading to hedonic or aversive responses, throughout memory updating. Finally, we will discuss evidence regarding memory updating and its potential clinical implication in drug addiction, phobias, and post-traumatic stress disorder.”
#Xiao X, Dong Q, Gao J, Men W, Poldrack RA, Xue G. Transformed Neural Pattern Reinstatement during Episodic Memory Retrieval. J Neurosci. 2017
https://pmc.ncbi.nlm.nih.gov/articles/PMC6596730/
Quote: “Contemporary models of episodic memory posit that remembering involves the reenactment of encoding processes. Although encoding-retrieval similarity has been consistently reported and linked to memory success, the nature of neural pattern reinstatement is poorly understood. Using high-resolution fMRI on human subjects, our results obtained clear evidence for item-specific pattern reinstatement in the frontoparietal cortex, even when the encoding-retrieval pairs shared no perceptual similarity. No item-specific pattern reinstatement was found in the ventral visual cortex. Importantly, the brain regions and voxels carrying item-specific representation differed significantly between encoding and retrieval, and the item specificity for encoding-retrieval similarity was smaller than that for encoding or retrieval, suggesting different nature of representations between encoding and retrieval. Moreover, cross-region representational similarity analysis suggests that the encoded representation in the ventral visual cortex was reinstated in the frontoparietal cortex during retrieval. Together, these results suggest that, in addition to reinstatement of the originally encoded pattern in the brain regions that perform encoding processes, retrieval may also involve the reinstatement of a transformed representation of the encoded information. These results emphasize the constructive nature of memory retrieval that helps to serve important adaptive functions.”
#Lee JLC, Nader K, Schiller D. An Update on Memory Reconsolidation Updating. Trends Cogn Sci. 2017
https://www.cell.com/trends/cognitive-sciences/abstract/S1364-6613(17)30078-5
Quote: “The reactivation of a synaptically stored memory in the brain can make the memory transiently labile. During the time it takes for the memory to re-stabilize (reconsolidate), the memory can either be reduced by an amnesic agent or enhanced by memory enhancers. The change in memory expression is related to changes in the brain correlates of long-term memory. Many have suggested that such retrieval-induced plasticity is ideally placed to enable memories to be updated with new information. This hypothesis has been tested experimentally, with a translational perspective, by attempts to update maladaptive memories in order to reduce their problematic impact. Here, we review the progress on reconsolidation-update studies, highlighting their translational exploitation and addressing recent challenges to the reconsolidation field.”
#Zhuang, L., Wang, J., Xiong, B. et al. Rapid neural reorganization during retrieval practice predicts subsequent long-term retention and false memory. Nat Hum Behav (2022).
https://doi.org/10.1038/s41562-021-01188-4
Quote: “Active retrieval can alter the strength and content of a memory, yielding either enhanced or distorted subsequent recall. However, how consolidation influences these retrieval-induced seemingly contradictory outcomes remains unknown. Here we show that rapid neural reorganization over an eight-run retrieval practice predicted subsequent recall. Retrieval practice boosted memory retention following a 24-hour (long-term) but not 30-minute delay, and increased false memory at both delays. Long-term retention gains were predicted by multi-voxel representation distinctiveness in the posterior parietal cortex (PPC) that increased progressively over retrieval practice. False memory was predicted by unstable representation distinctiveness in the medial temporal lobe (MTL). Retrieval practice enhanced the efficiency of memory-related brain networks, through building up PPC and MTL connections with the ventrolateral and dorsolateral prefrontal cortex that predicted long-term retention gains and false memory, respectively. Our findings indicate that retrieval-induced rapid neural reorganization together with consecutive consolidation fosters long-term retention and false memories via distinct pathways.”
#Karlsson Wirebring L, Wiklund-Hörnqvist C, Eriksson J, Andersson M, Jonsson B, Nyberg L. Lesser Neural Pattern Similarity across Repeated Tests Is Associated with Better Long-Term Memory Retention. J Neurosci. 2015
https://pmc.ncbi.nlm.nih.gov/articles/PMC6605150/
Quote: “Encoding and retrieval processes enhance long-term memory performance. The efficiency of encoding processes has recently been linked to representational consistency: the reactivation of a representation that gets more specific each time an item is further studied. Here we examined the complementary hypothesis of whether the efficiency of retrieval processes also is linked to representational consistency. Alternatively, recurrent retrieval might foster representational variability—the altering or adding of underlying memory representations. Human participants studied 60 Swahili–Swedish word pairs before being scanned with fMRI the same day and 1 week later. On Day 1, participants were tested three times on each word pair, and on Day 7 each pair was tested once. A BOLD signal change in right superior parietal cortex was associated with subsequent memory on Day 1 and with successful long-term retention on Day 7. A representational similarity analysis in this parietal region revealed that beneficial recurrent retrieval was associated with representational variability, such that the pattern similarity on Day 1 was lower for retrieved words subsequently remembered compared with those subsequently forgotten. This was mirrored by a monotonically decreased BOLD signal change in dorsolateral prefrontal cortex on Day 1 as a function of repeated successful retrieval for words subsequently remembered, but not for words subsequently forgotten. This reduction in prefrontal response could reflect reduced demands on cognitive control. Collectively, the results offer novel insights into why memory retention benefits from repeated retrieval, and they suggest fundamental differences between repeated study and repeated testing.”
– This also means that just because you remember something really well it does not mean your memory is correct. It just means that its assembly is really strong and vivid.
Vivid memories often feel more accurate than they actually are: vividness and accuracy are related but partially dissociable in humans. That is, vividness is typically neither necessary nor sufficient for accuracy.
#Windsor, P.M., Dering, B.R. and Donaldson, D.I. (2026) Does the Experience of Remembering Differentially Influence the Factual Accuracy of Recognition, and Confidence in Its Accuracy? Journal of Cognition
https://journalofcognition.org/articles/10.5334/joc.477
Quote: “There is ample evidence that vividness is associated with confidence in a memory’s accuracy. To the lay audience, for example, personal memories deemed vivid are more likely to be confidently believed. In addition, studies involving flashbulb memories (i.e., memories of historically significant world events) often report that such memories are detailed, long-lasting, and vivid (Brown & Kulik, 1977; Edery-Halpern & Nachson, 2004; Luminet & Spijkerman, 2017; Pillemer, Goldsmith, Panter & White, 1988). Vivid memories may not always be accurate, however, and other researchers have shown the contextual details of flashbulb memories can sometimes be incorrect or lacking (Stone, Luminet & Takahashi, 2015; Talarico, Kraha, Self & Boals, 2019; Talarico & Rubin, 2003; 2007). Equally, when asking whether remembering and its quality (i.e., vividness) might be one of many factors which underlie dissociations between confidence and accuracy, the current literature shows varying degrees of association between memory vividness and memory accuracy (e.g., Habermas & Diel, 2013; Richter, Cooper, Bays & Simons, 2016). Consequently, in addition to studying the experience of remembering and its influence on both confidence and memory accuracy, here we also assess how these relationships may be affected by variability in memory quality (i.e., remembered vividness).”
#Franziska R Richter, Rose A Cooper, Paul M Bays, Jon S Simons (2016) Distinct neural mechanisms underlie the success, precision, and vividness of episodic memory eLife
https://elifesciences.org/articles/18260
Quote: “A network of brain regions have been linked with episodic memory retrieval, but limited progress has been made in identifying the contributions of distinct parts of the network. Here, we utilized continuous measures of retrieval to dissociate three components of episodic memory: retrieval success, precision, and vividness. In the fMRI scanner, participants encoded objects that varied continuously on three features: color, orientation, and location. Participants’ memory was tested by having them recreate the appearance of the object features using a continuous dial, and continuous vividness judgments were recorded. Retrieval success, precision, and vividness were dissociable both behaviorally and neurally: successful versus unsuccessful retrieval was associated with hippocampal activity, retrieval precision scaled with activity in the angular gyrus, and vividness judgments tracked activity in the precuneus. The ability to dissociate these components of episodic memory reveals the benefit afforded by measuring memory on a continuous scale, allowing functional parcellation of the retrieval network.”
#Faul, L., Ritchey, M., & Kensinger, E. A. (2025). The relationship between subjective vividness and remembered visual characteristics of emotional stimuli across the lifespan. Emotion
https://psycnet.apa.org/doiLanding?doi=10.1037%2Femo0001518
Quote: “Episodic memories are characterized by the vividness of their recollection. Recent findings show that low-level visual properties can quickly fade from memory and that more vivid memories are associated with less fading. However, further work is needed to clarify this effect over longer delays and how it may shift based on the emotional valence of a stimulus, as well as one’s age. Here, participants (n = 307, aged 19–78, recruited in 2023–2024) incidentally encoded positive, negative, and neutral images shown at different levels of color saturation, contrast, and hue. At a next-day recognition test, images identified as old were rated on subjective vividness and then reconstructed based on the remembered visual information from encoding. More arousing images were recollected with more subjective vividness, and vividness ratings were primarily associated with biases in reconstructed color saturation, but in both instances, the coherence between these measures diminished with increasing age. Negative and neutral images showed memory fading (color saturation underestimations) at lower levels of subjective vividness, and neutral images additionally showed evidence of fading via contrast reconstruction. Positive images did not show evidence of fading and were reconstructed with inflated color saturation and contrast at all levels of vividness relative to negative and neutral images. Our findings show that subjective vividness is not uniformly related to remembered low-level visual information but differs depending on the visual information reconstructed, the emotionality of an experience, and individual differences such as age.”
– This is also why therapy can be so helpful. By revisiting hurtful memories in a safe context, ideally with helpful introspection, you are literally changing your brain. Literally rewiring yourself to get a chance to be happier.
Various forms of therapy are considered to be highly beneficial for people dealing with painful memories, such as those with post-traumatic stress disorder, though not everyone fully recovers.
Therapy that works with painful or traumatic memories might even be able to physically modify how those memories are stored and processed in brain networks for threat, emotion, and meaning. However, more research in this regard is needed, as most of these physical effects are not yet well understood.
#Watkins LE, Sprang KR and Rothbaum BO. Treating PTSD: A Review of Evidence-Based Psychotherapy Interventions. Front. Behav. Neurosci. (2018)
https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2018.00258/full
Quote: “Posttraumatic stress disorder (PTSD) is a chronic, often debilitating mental health disorder that may develop after a traumatic life event. Fortunately, effective psychological treatments for PTSD exist. In 2017, the Veterans Health Administration and Department of Defense (VA/DoD) and the American Psychological Association (APA) each published treatment guidelines for PTSD, which are a set of recommendations for providers who treat individuals with PTSD. The purpose of the current review article is to briefly review the methodology used in each set of 2017 guidelines and then discuss the psychological treatments of PTSD for adults that were strongly recommended by both sets of guidelines. Both guidelines strongly recommended use of Prolonged Exposure (PE), Cognitive Processing Therapy (CPT) and trauma-focused Cognitive Behavioral Therapy (CBT). Each of these treatments has a large evidence base and is trauma-focused, which means they directly address memories of the traumatic event or thoughts and feelings related to the traumatic event. Finally, we will discuss implications and future directions.”
#Foa EB, McLean CP. The Efficacy of Exposure Therapy for Anxiety-Related Disorders and Its Underlying Mechanisms: The Case of OCD and PTSD. Annu Rev Clin Psychol. 2016
https://www.annualreviews.org/content/journals/10.1146/annurev-clinpsy-021815-093533
Quote: “EXPOSURE THERAPY AND ITS THEORETICAL FOUNDATION
Exposure therapy is a set of treatment approaches that are frequently used to reduce the pathological fear and related emotions, such as guilt, that are common in anxiety-related disorders. During exposure, patients are encouraged to approach feared, but safe, objects, situations, thoughts, sensations, and memories, with the ultimate goal of reducing fear reactions to those stimuli. Exposure procedures are divided into three primary types: in vivo (real-life) exposure, imaginal exposure (revisiting the distressing traumatic memory in imagination), and interoceptive exposure. The selection of the type of exposure is determined by the pathological characteristics of a given disorder. Often, several types of exposure are used concurrently in exposure therapy programs.”
#Aarts, I., Thorsen, A.L., Vriend, C. et al. Effects of psychotherapy on brain activation during negative emotional processing in patients with posttraumatic stress disorder: a systematic review and meta-analysis. Brain Imaging and Behavior (2024)
https://doi.org/10.1007/s11682-023-00831-0
Quote: “Post-traumatic stress disorder (PTSD) is a debilitating condition which has been related to problems in emotional regulation, memory and cognitive control. Psychotherapy has a non-response rate of around 50% and understanding the neurobiological working mechanisms might help improve treatment. To integrate findings from multiple smaller studies, we performed the first meta-analysis of changes in brain activation with a specific focus on emotional processing after psychotherapy in PTSD patients. We performed a meta-analysis of brain activation changes after treatment during emotional processing for PTSD with seed-based d mapping using a pre-registered protocol (PROSPERO CRD42020211039). We analyzed twelve studies with 191 PTSD patients after screening 3700 studies. We performed systematic quality assessment both for the therapeutic interventions and neuroimaging methods. Analyses were done in the full sample and in a subset of studies that reported whole-brain results. We found decreased activation after psychotherapy in the left amygdala, (para)hippocampus, medial temporal lobe, inferior frontal gyrus, ventrolateral prefrontal cortex, right pallidum, anterior cingulate cortex, bilateral putamen, and insula. Decreased activation in the left amygdala and left ventrolateral PFC was also found in eight studies that reported whole-brain findings. Results did not survive correction for multiple comparisons. There is tentative support for decreased activation in the fear and cognitive control networks during emotional processing after psychotherapy for PTSD. Future studies would benefit from adopting a larger sample size, using designs that control for confounding variables, and investigating heterogeneity in symptom profiles and treatment response.”
Google Sites
Report abuse