# Frontiers in Neural Information Processing: Encoding, Memory, Navigation, and Fundamental Mechanisms Introduction The quest to understand how the intricate biological machinery of the brain gives rise to the richness of cognition and behavior remains one of the most profound challenges in science. Traditional neuroscience has made significant strides by focusing on established principles, such as the central role of neuronal action potentials (spikes) in information transmission and the primacy of synaptic plasticity in memory storage. However, emerging evidence and theoretical considerations compel us to look beyond these paradigms. This report synthesizes the current research landscape surrounding eight critical research questions that probe these frontiers, pushing the boundaries of our understanding of neural encoding, memory mechanisms, animal navigation, and the fundamental timescales and computational principles governing brain function. These questions challenge conventional assumptions: exploring the potential informational roles of non-spike neural signals like local field potentials, glial cell activity, and ephaptic fields; investigating whether memory storage relies on intracellular molecular mechanisms potentially independent of synaptic strength and persists through global synaptic resetting; delving into the neural and computational underpinnings of subjective memory recall and its inherent reconstructive fallibility; examining the sophisticated multi-sensory integration underlying complex behaviors like ant navigation beyond reliance on chemical trails; and probing the very limits of neural processing timescales, including the speculative role of sub-millisecond or even quantum phenomena, alongside the mechanisms bridging continuous sensory integration with discrete perceptual and motor events. Addressing these questions necessitates not only sophisticated experimental designs but also the development and application of advanced technologies. Multi-modal recording techniques, capable of simultaneously capturing electrical, optical, and chemical signals from neural tissue 1, and powerful computational models are becoming indispensable tools for dissecting the complex interplay of diverse signals and network states.8 This report aims to critically evaluate the evidence pertaining to these frontier questions, drawing upon recent experimental findings and theoretical frameworks, to provide a comprehensive overview of the current state of knowledge, identify key challenges, and illuminate the potential implications of future discoveries in these exciting and rapidly evolving domains of neuroscience. I. Neural Encoding and Information Transmission Beyond Spikes The dominant paradigm in neural coding has long centered on the timing and rate of action potentials. However, a growing body of research investigates the potential contributions of other physiological signals—local field potentials, glial cell activity, and ephaptic coupling—to information processing, questioning whether they carry unique information or merely reflect epiphenomena of spiking activity. A. The Informational Contribution of Non-Spike Signals (RQ1 Analysis) The investigation into whether non-spike signals carry unique sensory or cognitive information, non-redundant with concurrent neuronal spiking, requires dissecting the nature and content of these signals. - Local Field Potentials (LFPs): LFPs represent aggregate electrical activity, primarily reflecting summed postsynaptic potentials, afterpotentials, and other transmembrane currents from populations of neurons.13 Studies, particularly in motor cortex, demonstrate that LFPs contain significant information about movement parameters like direction and velocity.16 A key question is the degree of redundancy with spiking activity. High-frequency LFP components (e.g., >70-100 Hz) often show strong correlations with local spiking activity in terms of information content, potentially because these bands capture low-pass filtered action potentials from nearby neurons.14 However, lower frequency LFP bands (e.g., <30 Hz) exhibit increasing dissimilarity in directional tuning compared to spikes, suggesting they might reflect distinct neural processes, such as broader synaptic integration or network oscillations, rather than just filtered spiking.14 Indeed, LFP signals can show significant task-related modulation even when concurrent spiking activity on the same electrode does not.16 From a practical standpoint, LFPs hold promise for brain-computer interfaces (BCIs) due to their potential for greater long-term stability compared to single-unit spikes, although evidence suggests their decoding performance may degrade at similar rates over extended periods.16 The relationship between LFPs and underlying population dynamics is complex and context-dependent. LFP power across different frequency bands correlates with large-scale population activity patterns ('latent dynamics') in ways that vary across brain regions (e.g., M1 vs. PMd) and depend on frequency.13 This relationship appears stable across different phases of a task (planning vs. execution) but weakens during rest, indicating that LFPs reflect behaviorally relevant network states.13 A significant methodological challenge in interpreting LFP data is the contamination by spike waveforms, especially at frequencies above ~25-100 Hz.14 This "bleed-through" can create artifactual correlations between LFP power/phase and spiking activity. While methods exist to remove spike artifacts, they must be applied carefully to avoid distorting the genuine LFP signal, particularly its phase information, which itself can be informative.14 Furthermore, the assumption of spatial homogeneity of LFPs is often invalid, adding complexity to relating LFP recordings to specific neuronal populations.15 - Glial Cell Activity (Calcium Waves): The traditional view of glial cells, particularly astrocytes, as merely providing passive support for neurons has been overturned. Astrocytes are now recognized as active participants in neural circuits, forming "tripartite synapses" with pre- and postsynaptic neurons.17 They possess receptors for neurotransmitters and respond to neuronal activity with elevations in intracellular calcium (Ca2+).18 This Ca2+ signaling constitutes a form of astrocyte excitability.23 These signals can propagate within individual astrocytes and spread to neighboring astrocytes through gap junctions or via the release of chemical messengers ("gliotransmitters"), forming intercellular Ca2+ waves.17 Computational models propose that information might be encoded within these waves through amplitude modulation (AM) or frequency modulation (FM), potentially allowing for the integration of information across the astrocyte network.17 Functionally, astrocyte Ca2+ elevations trigger the release of gliotransmitters like glutamate, ATP (and its derivative adenosine), and D-serine.17 These gliotransmitters act back on neurons to modulate synaptic transmission (both excitatory and inhibitory), influence synaptic plasticity (LTP/LTD), and regulate neuronal excitability.17 This bidirectional communication implies that astrocytes actively participate in information processing within neuron-glia networks.18 Theoretical models have linked astrocyte activity and tripartite synapses to higher cognitive functions, including learning, memory consolidation, sensory integration, and even consciousness.17 Crucially, recent experimental work combining two-photon calcium imaging of neurons and astrocytes in awake, behaving mice demonstrated that astrocyte Ca2+ signals in hippocampal microdomains encode spatial information (animal's location) that is complementary and synergistic to the information encoded by nearby neuronal place cells.29 This provides direct evidence for astrocytes carrying unique spatial information. Astrocyte Ca2+ signals are spatially complex, occurring frequently in fine processes and microdomains near synapses, suggesting highly localized modulation of synaptic function.28 Studying these signals in vivo, particularly in humans, remains technically challenging 17, and further work is needed to decipher the specific coding schemes and definitively link glial Ca2+ dynamics to specific cognitive computations.28 - Ephaptic Coupling: This refers to the modulation of neuronal activity via extracellular electrical fields generated by neighboring neurons, a form of communication distinct from chemical or electrical synapses.34 The transmembrane currents associated with neuronal activity, particularly the large currents during action potential (AP) generation and propagation, create potential gradients in the extracellular space.40 These fields can directly influence the membrane potential of nearby neurons. AP annihilation at axon terminals is considered a particularly potent source of ephaptic fields.34 The strength and nature (excitatory or inhibitory) of ephaptic coupling depend critically on factors like the distance between neurons, their relative orientation, cell morphology, and the conductivity of the extracellular medium.34 Functional roles for ephaptic coupling have been documented in specific systems like the Mauthner cell circuit in fish and cerebellar Purkinje cells.34 Proposed functions include mediating ultra-fast interactions (faster than chemical synapses), promoting neuronal synchrony (especially in the gamma band), modulating network excitability, and potentially contributing to information processing.35 Recent work suggests ephaptic coupling might play a role in shaping memory engram complexes by guiding ensemble activity.37 Modeling studies also indicate that ephaptic interactions within white matter tracts could influence signal propagation speed and lead to complex spatiotemporal patterns like phase locking or traveling fronts.36 While often viewed as merely modulatory, computational models suggest that the specific spatiotemporal profile of ephaptic fields, particularly those generated by AP annihilation at terminals, can induce distinct temporal patterns of hyperpolarization and depolarization in target neurons, potentially conveying more structured information than simple excitability changes.34 Supporting this, evidence from primate prefrontal cortex suggests that electric fields (potentially reflecting ephaptic interactions) carry information about working memory contents and are more stable over time than the underlying spiking activity.37 Experiments show single Purkinje cell spikes can generate fields sufficient to influence neighbors 39, and synchronous volleys can generate fields large enough to modulate axonal conduction velocity.40 However, experimentally isolating and quantifying ephaptic effects in vivo remains a major challenge, requiring sophisticated modeling approaches (e.g., finite element modeling, realistic cable models) and careful experimental design to distinguish ephaptic influences from synaptic and other interactions.34 - Synthesis: The available evidence strongly indicates that non-spike signals are not mere echoes of neuronal firing. LFPs reflect integrated synaptic inputs and network state 13, glial calcium signals represent slower integrative processes and neuromodulation with unique spatial encoding capabilities 17, and ephaptic fields mediate ultra-fast interactions sensitive to geometry and synchrony.34 Their distinct biophysical origins, spatiotemporal scales, and demonstrated information content (e.g., LFP decoding 16, glial spatial coding 29, ephaptic memory coding 37) point towards complementary information streams. A comprehensive understanding of neural computation thus necessitates a shift away from a purely spike-centric view towards one that integrates information across these multiple signaling modalities. This, however, demands advanced multi-modal recording technologies 1 and rigorous analytical methods that carefully account for potential confounds, such as spike contamination of LFPs 14, the complex dynamics of glial signals 30, and the biophysical determinants of ephaptic coupling.34 - Table 1: Comparison of Non-Spike Neural Signals | | | | | | | | |---|---|---|---|---|---|---| |Signal Type|Primary Origin|Temporal Scale|Spatial Scale|Information Content (vs. Spikes)|Stability (vs. Spikes)|Key Recording/Analysis Challenges| |LFP (Low-freq <70Hz)|Summed postsynaptic potentials, other transmembrane currents, network oscillations 13|Tens to hundreds of ms|Mesoscopic (hundreds of µm to mm)|Partially distinct/complementary; reflects input/network state more than output 14|Comparable degradation, potentially higher intersession stability in some cases 16|Spike bleed-through, spatial inhomogeneity, relating signal to specific populations 14| |LFP (High-freq >70Hz)|Summed potentials, often includes filtered spikes 14|Milliseconds to tens of ms|Mesoscopic (potentially more local than low-freq)|Often highly correlated/redundant with local spiking 16|Similar degradation rates 16|Strong spike bleed-through artifact risk 14| |Glial Ca++ Waves|Intracellular Ca2+ release/influx in astrocytes, triggered by neuronal activity/neuromodulators 18|Seconds to minutes|Single cell to syncytium (µm to mm)|Complementary/synergistic; encodes spatial info, network state, modulates plasticity 17|Likely high (reflects slower processes), but long-term studies needed|Imaging depth/resolution, distinguishing glial vs. neuronal signals, interpreting complex spatiotemporal dynamics 17| |Ephaptic Fields|Extracellular potentials from transmembrane currents (esp. APs, AP annihilation) 34|Sub-millisecond to milliseconds|Microscopic to mesoscopic (highly geometry-dependent)|Potentially distinct; reflects geometry/synchrony, may mediate fast interactions, linked to memory content 34|Potentially high (fields linked to stable structures), but depends on underlying activity stability 37|Isolating from synaptic effects, accurate biophysical modeling, in vivo measurement difficulty 34| B. Network Context and the Nature of the Neural Code (RQ2 Analysis) The traditional approach to neural decoding often assumes a relatively stable mapping between neural activity (typically spike rates or patterns) and the variable being decoded (e.g., a stimulus or movement). However, accumulating evidence suggests that neural representations are highly dynamic and sensitive to the broader network context, encompassing factors like neuromodulatory state, ongoing oscillations, the activity of non-neuronal cells, and task demands. This raises fundamental questions about the nature of the neural code and whether variability often treated as "noise" might actually represent context-dependent signals. - Context-Dependent Neural Representations: Neural responses are not fixed but are actively modulated by the context in which they occur. For example, in the auditory cortex, the representation of a sound can be altered by the animal's expectation of reward associated with that sound, even though the sound's identity remains decodable.44 Similarly, fMRI studies show that anticipatory neural activity patterns preparing for a specific stimulus category (e.g., face vs. name) differ depending on whether the preparation is driven by attentional relevance or probabilistic expectation, indicating context shapes preparatory states.45 This context sensitivity implies that a simple, static decoder may fail to capture the true relationship between neural activity and the encoded variable across different situations. - Limitations of Simple Decoders and the Rise of Non-Linear Models: The dynamic nature of neural coding often necessitates more sophisticated decoding approaches than simple linear models. For instance, during development in the rat motor cortex (M1), the relationship between neural activity and limb movements transitions from being linearly decodable during early, discontinuous activity phases to requiring non-linear decoders (like Long Short-Term Memory networks, LSTMs) after the emergence of continuous cortical activity.9 This shift suggests that the underlying neural code becomes more complex, potentially incorporating temporal dependencies or population interactions that linear models cannot capture. Across various domains, deep learning models are increasingly outperforming traditional linear or simpler machine learning approaches in neural decoding tasks, particularly when dealing with complex signals like EEG or LFP, or when the underlying coding scheme is unknown or non-linear.8 - Reinterpreting "Noise": The Role of Correlations and Latent Dynamics: Neural responses exhibit variability even to repeated presentations of the same stimulus. This variability, often termed "noise," is frequently correlated across neurons within a population ("noise correlations").47 Ignoring these correlations during decoding can lead to a significant loss of information.47 The impact of noise correlations on information coding is complex and depends critically on the relationship between the correlation structure and the structure of the signal (i.e., how neurons are tuned to the stimulus).47 What appears as noise relative to the stimulus might reflect fluctuations in unmeasured contextual variables, such as attention, arousal, or internal state, that are shared across the population. Theoretical work suggests that even small correlations can be information-limiting in large populations, especially if they align with the signal direction.48 Decoding frameworks that explicitly model this correlated variability, for instance, by identifying and removing task-irrelevant low-dimensional latent dynamics ("denoising"), can improve decoding performance.55 Furthermore, ubiquitous computational mechanisms like divisive normalization are known to shape noise correlation structures.52 This suggests that much of the "noise" is, in fact, structured variability reflecting network state and context. - Network State: Oscillations, Neuromodulation, and Glial Activity: The overall state of the brain network profoundly influences information processing and coding. Brain states associated with sleep, anesthesia, or wakefulness exhibit distinct oscillatory patterns and levels of synchrony, impacting information capacity.56 Neuromodulatory systems (e.g., cholinergic, noradrenergic, dopaminergic) dynamically regulate network oscillations, including cross-frequency coupling (CFC), which is thought to be crucial for coordinating activity across brain regions and supporting cognitive functions.57 Incorporating oscillatory features (from LFP/EEG) into decoding models is a promising avenue.8 Beyond neuronal factors, the state of glial networks, particularly astrocyte activity, also contributes to the network context by modulating synaptic transmission and potentially encoding complementary information.17 - Context-Aware Computational Models: Recognizing the importance of context, researchers are developing computational models that explicitly incorporate it. For instance, in speech decoding, using context-dependent units like diphones (capturing transitions between phonemes) rather than context-independent phonemes improves decoding accuracy from neural signals, suggesting neural representations encode contextual dependencies.58 Generative models 10 and latent variable approaches 55 aim to capture the underlying structure of neural activity, including context-dependent variations, rather than just fitting a direct mapping. Formalizing context-dependent learning under uncertainty is also a key area of theoretical development.64 Comparing different model architectures (e.g., psychologically plausible models vs. large neural language models) on diverse, multi-modal datasets is essential for identifying which computational principles best capture context-dependent neural processing.10 - Synthesis: The evidence strongly suggests that neural codes are not static input-output mappings but are dynamically shaped by the network's internal state and the broader behavioral context.9 This dynamic nature necessitates decoding approaches that move beyond simple linear assumptions and account for context.9 Much of the response variability previously dismissed as noise likely contains meaningful signals related to these contextual factors, such as attention, neuromodulation, or ongoing network oscillations.47 Ignoring this structure, particularly noise correlations, leads to suboptimal decoding. Progress in understanding the neural code therefore depends critically on the synergistic development of multi-modal experimental techniques capable of capturing diverse aspects of network context (e.g., LFPs, neuromodulators, glial activity, behavior) 1 and sophisticated computational models (e.g., deep learning, state-space models, generative approaches) that can learn and represent these complex, dynamic, context-dependent relationships.8 - Table 2: Comparison of Neural Decoding Model Approaches | | | | | | |---|---|---|---|---| |Model Class|Key Assumption(s)|Handling of Context/Noise|Supporting Evidence/Snippets|Limitations/Challenges| |Linear Models|Linear relationship between features (e.g., spike rates) and decoded variable. Static code.|Typically ignores context. Treats variability as noise.|Sufficient for some tasks/states (e.g., early M1 dev. 9).|Fails when coding is non-linear or context-dependent.9 Ignores informative correlations.47| |Deep Learning (CNN/RNN/LSTM)|Can learn complex, non-linear mappings. Can model temporal dependencies (RNN/LSTM).|Can implicitly capture context/temporal structure from data. May still treat residual variability as noise unless specifically designed otherwise.|Outperform linear models in many cases, esp. with complex/dynamic data.8|Require large datasets. Can be "black boxes," harder to interpret mechanisms. May overfit if not regularized.| |Latent Variable Models|Observed activity reflects lower-dimensional underlying dynamics + noise. Correlated variability is structured.|Explicitly models shared variability (latent dynamics). Can separate task-relevant dimensions from "noise" dimensions (denoising).|Can improve decoding by removing task-irrelevant variability.55 Aligns with observations of low-D population dynamics.13|Requires assumptions about latent structure. Estimating latent variables can be challenging.| |Context-Explicit Models (e.g., DCoND, Generative Models)|Neural representations explicitly depend on context. Generative models capture underlying causal structure.|Directly incorporate context variables or dependencies into the model structure (e.g., diphones 58, task rules 64, priors 66). Aim to model the generative process including noise/context.|Improved speech decoding with context-dependent units.58 Frameworks for context-dependent learning.64 Generative models fit complex data.66|Require rich data capturing context. Model specification can be complex. Inference in generative models can be difficult.| II. Mechanisms of Memory Storage and Recall Memory research is moving beyond purely synaptic explanations to explore the roles of intracellular molecular processes, non-synaptic plasticity, and the computational mechanisms underlying the subjective experience and reconstructive nature of recall. A. Intracellular Molecular Basis of the Engram (RQ3 Analysis) The concept of the memory engram posits that specific neuronal ensembles, activated during learning, undergo persistent changes to store a memory trace.68 A critical question is whether manipulating molecular components within these engram cells can alter the memory itself, potentially independently of changes in the synaptic connections between these cells. - Identifying and Accessing Engram Cells: Significant advances have been made in "engram tagging" technologies. These methods typically leverage the promoters of activity-dependent immediate-early genes (IEGs), such as c-fos or Arc, which are transiently upregulated in active neurons. By combining these promoters with genetic tools (e.g., tetracycline-controlled systems like TetTag, Cre-recombinase systems like TRAP, or viral vectors) researchers can permanently label or express specific proteins (reporters like GFP, or manipulators like channelrhodopsin (ChR2), DREADDs (hM3Dq/hM4Di), or dominant-negative proteins like mCREB) specifically in the population of neurons activated during a particular learning event.69 These techniques allow for the visualization, activation, inhibition, or molecular manipulation of putative engram ensembles.69 - Intracellular Molecular Mechanisms: Several molecular pathways within engram cells are implicated in memory storage: - Epigenetics: Modifications to chromatin structure, such as histone acetylation and DNA methylation/hydroxymethylation, are proposed as key mechanisms for long-term information storage within engram cells.78 The histone acetyltransferase CREB-binding protein (CBP) is essential for memory, and its disruption causes cognitive deficits, while inhibiting histone deacetylases (HDACs) can enhance memory, particularly when combined with neural activity (epigenetic priming).78 - Transcription Factors: The transcription factor CREB plays a pivotal role. Neurons with elevated CREB levels are preferentially allocated to fear or drug-associated memory engrams.79 Selectively disrupting CREB function (using EGFP-mCREB expression driven by the Fos promoter) within dentate gyrus (DG) engram cells after fear conditioning impairs the consolidation of long-term contextual fear memory.76 This manipulation points to CREB-dependent transcription within the engram as crucial for stabilizing the memory trace. CREB regulates downstream genes involved in synaptic plasticity.76 - RNA and Proteins: While standard consolidation and reconsolidation processes require RNA and protein synthesis 80, specific molecules within engram cells show persistent changes. The IEG protein Arc, for example, exhibits sustained expression in DG engram cells 24 hours after fear conditioning, enabling transcriptomic analysis of these cells during consolidation.76 The protein Tau, particularly its phosphorylation state at residue T205 within engram cells during encoding, has been shown to be necessary for the efficient retrieval of remote memories via natural cues, although not for recall elicited by direct optogenetic activation of the engram.84 This suggests Tau acts as a "gatekeeper" linking the engram state to retrieval pathways. - Targeted Molecular Manipulation: Optogenetic and chemogenetic tools are widely used to demonstrate the necessity and sufficiency of engram cell activity for memory recall.69 These can be combined with molecular manipulations, such as expressing dominant-negative CREB.76 More recently, CRISPR-Cas9 gene editing technology offers the potential for precise, conditional genetic modifications within tagged engram cells.85 Studies have used CRISPR to knock down cbp in engram cells, resulting in reduced dendritic spine density and impaired memory recall 85, or to target genes in specific engram projections relevant to extinction learning.85 - The Challenge of Dissociation: The core difficulty posed by RQ3 is demonstrating that intracellular molecular manipulations alter memory independently of any changes in synaptic strength or function. While some evidence hints at such a dissociation – for example, memories seemingly retained despite lack of synaptic reinforcement 85, recall via direct engram activation despite failed natural recall due to Tau manipulation 84, or recall persistence after consolidation blockade 78 – definitively proving this independence is challenging. Many of the molecules implicated (CREB, CBP, Arc, Tau) are known regulators of synaptic plasticity and structure.76 Therefore, manipulating them intracellularly is highly likely to have downstream consequences at the synapse, even if not immediately apparent or measured. Rigorous controls measuring synaptic function concurrently with molecular manipulation are essential but difficult. Future studies might target intracellular molecules with no known direct role in synaptic structure or transmission. - Synthesis: Memory engrams are not static entities defined solely by strengthened synapses. They undergo plasticity at multiple levels, including intracellular molecular and epigenetic changes.76 These intracellular modifications, potentially involving epigenetic priming 78, may establish a latent memory trace that contributes to the persistence and potential reactivation of the memory, complementing changes in synaptic connectivity.78 The ability to probe these intracellular mechanisms is advancing rapidly due to powerful engram-tagging and molecular manipulation tools like CRISPR.69 However, conclusively demonstrating memory storage independent of synaptic efficacy remains a significant experimental hurdle, requiring innovative approaches to disentangle intracellular state changes from their inevitable influence on synaptic function. B. Memory Persistence Beyond Synaptic Strength Normalization (RQ4 Analysis) The traditional view posits that memories are stored through long-lasting changes in synaptic strength. However, the brain exhibits homeostatic mechanisms, particularly during sleep, that appear to globally normalize or downscale synaptic weights. This raises a fundamental question: how do specific memories persist through such global normalization processes? Do they remain encoded, perhaps latently, and if so, what mechanisms enable their stability and subsequent retrieval? - Synaptic Homeostasis and Sleep: The Synaptic Homeostasis Hypothesis (SHY) proposes that a key function of sleep, particularly slow-wave sleep (SWS), is to counteract the net increase in synaptic potentiation that occurs during wakefulness.86 This downscaling is thought to prevent synaptic saturation, conserve energy, and restore the capacity for learning new information upon waking.86 Evidence supporting sleep-dependent synaptic downscaling includes observations of reduced AMPA receptor density and dendritic spine size after sleep compared to wakefulness, and the restoration of cortical plasticity capacity by sleep.86 This homeostatic regulation involves negative feedback mechanisms operating at synaptic and neuronal levels.87 - Sleep, Oscillations, and Memory Consolidation: Paradoxically, sleep, especially SWS, is also crucial for memory consolidation.91 During SWS, characterized by slow oscillations (<1-4 Hz) and sleep spindles, newly encoded memories are thought to be reactivated and redistributed from temporary storage sites like the hippocampus to more permanent cortical networks (active system consolidation).86 This process involves the replay of neuronal firing patterns associated with prior learning experiences, often coordinated by hippocampal sharp-wave ripples nested within cortical slow oscillations and thalamocortical spindles.91 - Reconciling Downscaling and Consolidation: The apparent conflict between global synaptic downscaling and selective memory consolidation during sleep suggests several possibilities: - Selective Synaptic Sparing: Synaptic downscaling may not be uniform. SHY proposes a competitive process where stronger, more salient synapses, potentially those tagged during learning as part of an engram, are preferentially spared or even strengthened, while weaker, less relevant synapses are pruned.86 Memory replay during sleep oscillations could serve as the mechanism for identifying and protecting these important synapses.86 Computational models suggest sleep replay can reorganize synaptic connectivity to accommodate both old and new memories, minimizing interference.93 - Non-Synaptic Storage Mechanisms: The persistence of memory information through periods of potential synaptic weight normalization strongly points towards the involvement of storage mechanisms beyond synaptic efficacy alone. These could provide a more stable substrate for long-term memory: - Intracellular/Epigenetic States: As discussed for RQ3, stable molecular or epigenetic modifications within engram neurons (e.g., histone modifications, DNA methylation) could encode memory information in a latent form, resistant to fluctuations in synaptic weights.78 - Nonsynaptic Plasticity (Intrinsic Excitability): Long-term changes in the function, density, or distribution of ion channels in the soma, dendrites, or axon can alter a neuron's intrinsic excitability and firing properties.87 Such changes could make engram neurons more likely to be reactivated in specific patterns, effectively storing information about past activity, even if individual synaptic weights are globally scaled.94 - Structural Plasticity: More permanent changes, such as the formation or elimination of synapses, or the remodeling of dendritic or axonal structures, could provide a stable physical basis for memory that persists longer than specific weight values.88 Sleep is known to promote protein synthesis, which could support such structural rearrangements.91 - Glial Contributions: Given their role in modulating synaptic function and network activity (RQ1), astrocytes or other glial cells might contribute to maintaining network states or synaptic configurations over long periods. - Evidence for Latent Memory Persistence: Studies on memory reconsolidation provide indirect evidence for latent memory traces. When a consolidated memory is reactivated, it can become temporarily labile and susceptible to disruption (e.g., by protein synthesis inhibitors).80 However, such blockade often impairs the expression of the memory rather than completely erasing it, suggesting the trace persists in an inaccessible or latent state.80 The ability to pharmacologically (e.g., HDAC inhibitors 78) or optogenetically 84 reinstate memories that were previously unrecallable further supports the existence of dormant or latent engrams. - Synthesis: The phenomenon of memory consolidation during sleep, despite concurrent homeostatic synaptic downscaling, strongly challenges the notion that memories are stored solely as static synaptic weights. Persistence through global normalization implies that memory information is likely encoded through multiple, interacting mechanisms.80 Sleep appears to be a critical period not just for passive synaptic renormalization but for active, selective processes that reorganize memory traces. Replay mechanisms likely interact with homeostatic pressures to preserve salient information while pruning weaker connections, potentially by strengthening specific engrams or tagging them for protection.86 Crucially, non-synaptic mechanisms, including stable intracellular molecular states (epigenetics), long-term changes in intrinsic neuronal excitability, and structural plasticity, are emerging as essential contributors to the enduring nature of memory.78 Investigating how these diverse forms of plasticity interact, particularly during offline states like sleep, is vital for a complete understanding of long-term memory storage and retrieval. C. Neural and Computational Basis of Recall Qualia and Reconstruction (RQ5 Analysis) Memory recall is not a simple playback of past events but a dynamic, reconstructive process imbued with subjective qualities, or "qualia," such as vividness and a feeling of re-experiencing. This reconstruction is imperfect, leading to errors, distortions, and even confabulations. Understanding the neural dynamics that differentiate vivid recall from imagination or simple recognition, and developing computational models that capture the reconstructive nature of memory, are key challenges. - Neural Correlates of Recall, Imagination, and Vividness: Neuroimaging studies reveal both overlap and divergence in the brain networks supporting remembering the past versus imagining the future or fictitious events. - Shared Network: A core network including the medial temporal lobe (MTL, especially hippocampus), medial prefrontal cortex (mPFC), posterior cingulate cortex/retrosplenial cortex (PCC/RSC), and lateral parietal and temporal regions is commonly activated during both episodic memory retrieval and imagination/simulation of future events.96 This overlap underpins the "constructive episodic simulation" hypothesis, suggesting that memory provides the building blocks for imagination.96 - Differential Activation: Subtle but consistent differences exist. Imagining novel future events often elicits stronger activation in anterior hippocampal and frontopolar regions, possibly reflecting greater demands on constructive processes and recombination of information.96 Conversely, remembering actual past events tends to activate posterior regions, including visual cortex and posterior hippocampus/parahippocampal gyrus, more strongly, likely reflecting the reactivation of sensory and perceptual details from the original experience.96 Anterior mPFC and PCC activity can also be greater during retrieval of actual versus fictitious events.96 - Vividness and Connectivity: The subjective vividness of recall correlates with hippocampal activity.100 High-resolution fMRI reveals engagement of all hippocampal subfields during vivid autobiographical memory recall, with the anterior body of the pre/parasubiculum showing unique engagement and strong functional connectivity to the core recall network (vmPFC, parietal regions).101 Interestingly, the nature of vividness matters: recalling internal details (thoughts, feelings) is associated with increased hippocampus-mPFC connectivity, whereas recalling external perceptual details is associated with decreased connectivity in the same pathway.100 This suggests distinct neural interactions support different facets of subjective experience during recall. Temporal factors also play a role, with recent memories often eliciting stronger hippocampal activation and connectivity than remote ones.98 - Temporal Dynamics: Techniques like EEG and MEG provide the temporal resolution needed to investigate the dynamic interplay within these networks during recall and imagination, complementing the spatial information from fMRI.102 - Computational Models of Reconstruction and Error: Memory's reconstructive nature implies that recall involves actively generating a representation based on stored information and cues, rather than passively retrieving a fixed trace.99 This generative process is susceptible to errors. - Predictive Coding (PC): This framework views perception and potentially memory recall as processes of minimizing prediction errors based on internal generative models.106 Recall can be conceptualized as the brain generating predictions (simulating the past event) based on a retrieved memory trace (the internal model) and available cues.66 Errors, illusions, or confabulations can arise if strong priors (existing memories or beliefs) override conflicting cues or sensory details, or if the prediction error signals are noisy or misinterpreted.107 Predictive Coding Networks (PCNs) have been shown to act as effective associative memories, capable of reconstructing complete patterns from corrupted or highly incomplete cues (e.g., reconstructing ImageNet images from a small fraction of pixels).66 They achieve this through iterative error minimization within a hierarchical generative network. Dynamic PC models can learn and recall temporal sequences.110 - Generative Adversarial Networks (GANs): The GAN framework, involving a generator network creating samples and a discriminator network evaluating their realism, has been proposed as a model for implicit generative processes in the brain.67 The generator could potentially reconstruct memories, while the discriminator assesses their plausibility. Failures in this adversarial process, where internally generated reconstructions are misjudged as real, could offer a mechanism for phenomena like confabulation or hallucinations.67 - Confabulation Models: Confabulation, the unintentional generation of false or distorted memories, is often associated with damage to prefrontal regions involved in monitoring, temporal context processing, and reality filtering.114 Cognitive models emphasize deficits in strategic retrieval, source monitoring, or temporal consciousness.114 Computational approaches like Confabulation Theory propose specific architectures based on cortical modules and associative links to explain how retrieval cues can trigger inappropriate completions.117 The tendency of Large Language Models (LLMs) to generate plausible but factually incorrect information has been likened to confabulation, arising from pattern completion based on training data rather than veridical knowledge retrieval.119 - Testing Models: Evaluating and distinguishing these computational models requires linking their internal mechanisms and predictions to observable neural activity and behavior. Neuroimaging can identify correlates of model components (e.g., prediction error signals in PC 107). Brain stimulation techniques like Transcranial Magnetic Stimulation (TMS) or potentially optogenetics (in animal models) can be used to causally perturb specific brain regions or network dynamics implicated by the models and observe the effects on memory recall accuracy, vividness, or error rates.120 For example, TMS targeting parietal nodes of the episodic memory network has been shown to modulate recall performance and network connectivity.122 - Synthesis: The subjective quality of memory recall, its vividness and feeling of re-experiencing, appears tightly linked to the underlying neural processes of reconstruction. Vivid recall likely involves a dynamic interplay between brain regions responsible for retrieving stored associative information (MTL), regions involved in constructing coherent scenes or narratives (mPFC, parietal cortex), and regions responsible for reactivating the sensory details of the original experience (sensory cortices, posterior MTL).96 Differences between vivid recall and imagination may lie in the relative balance and specific content generated by these constructive versus reactivating processes. Computational frameworks like predictive coding and GANs offer powerful, biologically plausible mechanisms for understanding memory retrieval as an active, generative process.66 These models naturally account for the reconstructive nature of memory, including pattern completion from partial cues and the generation of errors or confabulations when the generative process is flawed or poorly constrained. Bridging the gap between subjective experience, neural dynamics, and computational theory requires integrating high-resolution neuroimaging (fMRI, EEG/MEG), detailed behavioral and phenomenological assessments, and causal methods like brain stimulation guided by model predictions.102 III. Sensory Integration in Animal Behavior: Ant Navigation Ants exhibit remarkable navigational capabilities, often relying on multiple sensory cues to find their way between the nest and resources. Understanding how they integrate information from different modalities, especially when dominant cues like pheromones are absent or unreliable, provides insights into the principles of robust biological navigation and the role of learning in shaping complex behaviors. A. Multi-Modal Navigation in Ants Beyond Pheromones (RQ6 Analysis) While pheromone trails are a well-known strategy for many ant species, particularly those engaging in mass recruitment, navigation in ants involves a much richer repertoire of sensory cues and computational strategies. This is especially true for solitary foragers or when pheromone trails are unavailable, ambiguous, or misleading. - The Ant Navigational Toolkit: Ants utilize a diverse set of cues: - Path Integration (PI): A fundamental mechanism where ants continuously track their net displacement from a starting point (usually the nest) by integrating estimates of direction and distance traveled. Direction is derived primarily from celestial compass cues (sun position, skylight polarization pattern) and potentially magnetic fields, while distance is estimated via an odometer (step counting, optic flow).128 PI allows ants to compute a direct "home vector" even after complex foraging paths. - Visual Cues: Ants learn and use visual information extensively. This includes recognizing panoramic scenes, using specific landmarks, matching skyline patterns, or utilizing canopy structure.128 Navigation often involves comparing the current view to stored visual memories, effectively using the panorama as a visual compass or employing snapshot-matching strategies.128 - Magnetic Field Sensing: There is growing evidence that ants, including desert ants (Cataglyphis) and leaf-cutter ants, can perceive the Earth's magnetic field.129 This sense can provide compass information and appears to be integrated into central brain structures like the central complex and mushroom bodies, influencing neuronal plasticity during visual learning.147 It might also serve as a stable reference system for calibrating other compasses.147 Ants can also learn artificial magnetic fields as positional landmarks.146 - Vibrational Cues: Ants are sensitive to substrate vibrations, primarily used for communication.146 However, experiments show that Cataglyphis noda can be trained to use localized, artificial vibrations as a landmark to pinpoint their nest entrance, demonstrating a capacity to learn this modality for navigation, even if its natural relevance as a landmark is uncertain.146 - Wind Direction: Particularly for desert ants navigating open terrain, wind direction can serve as a compass cue and influences foraging routes and learning walks.137 - Other Cues: Ants also utilize non-pheromone environmental odors as landmarks or plumes 130, tactile cues 146, potentially gravity 146 and thermal cues.146 When lost, many ants employ systematic search patterns or backtracking strategies.128 - Cue Integration, Hierarchy, and Flexibility: Ants rarely rely on a single cue but integrate information from multiple sources.131 This integration is dynamic and context-dependent. A hierarchy often exists, but it is flexible and modifiable by experience. - PI as a Scaffold: PI is crucial for initial exploration and provides a foundational vector.137 It acts as a "scaffold" upon which visual landmark information is learned and calibrated during learning walks.133 - Learned Cues Dominate: With experience, ants often prioritize learned cues, particularly visual landmarks, over their PI estimate or even pheromone trails if they conflict.130 Trunk-trail forming ants switch from odor trails (naïve) to visual landmarks (experienced).130 - Contextual Weighting: The relative weighting of different cues depends on their perceived reliability or certainty in a given context. Models propose optimal integration mechanisms where cues are weighted by their certainty.138 PI can also gate the response to other cues; for example, desert ants may only follow nest odor plumes when their PI indicates they are close to the nest.156 - Flexible Binding: Multimodal cues (e.g., visual + olfactory + airflow) can sometimes be learned and treated as a single configural unit ("bound"), such that removing one component impairs navigation. However, this binding is flexible and context-dependent; in simpler tasks, ants can navigate successfully using individual components of a previously learned multimodal stimulus.153 - Learning and Adaptation: Individual learning is paramount in ant navigation. - Learning Walks: Naïve ants perform structured learning walks around the nest to acquire visual panorama information, guided initially by PI and potentially magnetic cues.131 The structure of these walks is influenced by environmental factors like wind.137 - Associative Learning: Ants readily learn associations between environmental cues (visual, olfactory, magnetic, vibrational) and significant locations (nest, food) or outcomes (reward, punishment).146 - Modifying Innate Responses: Ants can learn to override innate responses. For example, Lasius niger can learn to ignore a conspecific pheromone trail if it consistently leads to punishment.166 However, this flexibility has limits; these ants struggle to learn to actively avoid the pheromone trail, suggesting constraints on reversing strong innate tendencies.166 - Experience-Dependent Reliance: As ants gain experience, their reliance on different cues shifts, typically towards learned visual information in familiar environments.130 - Sensory Ecology and Specialization: The specific navigational strategies employed are shaped by the ant's environment and morphology. Ants in visually cluttered environments or arboreal habitats may rely less on PI and more on visual or chemical cues, while those in open deserts rely heavily on PI and celestial cues.128 Species with larger eyes show a stronger preference for view-based navigation.128 - Synthesis: Ant navigation provides a compelling example of robust biological computation achieved through multi-modal sensory integration, flexible cue weighting, and adaptive learning. Rather than relying solely on pheromones, ants utilize a sophisticated toolkit including path integration, visual scene recognition, magnetic sensing, vibrational cues, and wind direction.128 The remarkable robustness of their navigation stems from this redundancy and their ability to flexibly integrate and weight cues based on context and reliability.138 Individual learning is crucial, allowing ants to acquire environmental knowledge (especially visual landmarks scaffolded by PI 133) and adapt their cue hierarchy, often prioritizing learned information over innate tendencies or less reliable cues like PI over long distances.130 This adaptive capability is further shaped by the ant's specific sensory ecology and morphology, leading to specialized navigational profiles across different species and habitats.128 IV. Fundamental Timescales and Mechanisms in Neural Processing Beyond specific functions like memory or navigation, fundamental questions remain about the basic operational principles and temporal limits of neural computation. Two such frontiers involve the potential role of ultra-fast or quantum processes and the mechanisms governing the transition between continuous information integration and discrete neural events. A. Exploring Sub-Millisecond and Quantum Neural Phenomena (RQ7 Analysis) Standard models of neural computation operate on timescales of milliseconds, dictated by membrane time constants, synaptic integration, and action potential duration. The possibility that functionally relevant processes occur at sub-millisecond timescales, or involve quantum mechanical effects, remains highly speculative but theoretically intriguing. - Sub-Millisecond Neural Processing: While action potentials themselves are brief (~1 ms), the question is whether distinct computational processes occurring within a sub-millisecond window significantly impact neuronal function. - Evidence for Precise Timing: There is compelling evidence that the precise timing of spikes, often at millisecond or even sub-millisecond resolution, carries significant information. In the fly visual system, spike timing variations on this scale encode distinct features of continuous motion stimuli, containing information not present in coarser time bins.168 Similarly, sub-millisecond synchronous firing among hippocampal pyramidal neurons is observed and correlates with behavioral task events, suggesting a role in effective information propagation.169 - Distinction from Sub-Millisecond Mechanisms: It is crucial to distinguish the precision of spike timing from the existence of underlying mechanisms operating at sub-millisecond speeds that influence computation. While precise timing is established, direct evidence for functionally relevant biophysical or computational processes occurring inherently faster than a millisecond and impacting subsequent neuronal firing or network dynamics is currently lacking in the provided materials. Much research on sub-second processing focuses on the perception and production of time intervals in the hundreds of milliseconds range, rather than fundamental sub-millisecond computations within neurons.170 Neural population codes can reconfigure rapidly, potentially on sub-trial timescales, but this reflects network flexibility rather than sub-millisecond computations within individual elements.173 - Quantum Effects in Neural Function: The potential role of quantum mechanics in brain function, particularly consciousness, is highly controversial. - Orchestrated Objective Reduction (Orch OR): This theory, proposed by Penrose and Hameroff, posits that consciousness arises from quantum computations occurring within microtubules, protein polymers inside neurons.174 Tubulin subunits within the microtubule lattice are proposed to act as quantum bits (qubits), capable of existing in quantum superposition. These quantum states are "orchestrated" by cellular processes and ultimately collapse via Penrose's proposed mechanism of "objective reduction" (OR), linked to quantum gravity, resulting in discrete moments of conscious experience.174 - Arguments and Putative Evidence: Proponents cite the discovery of quantum effects (like coherence) persisting at biological temperatures in other systems (e.g., photosynthesis) and recent reports of quantum vibrations detected in microtubules.174 They also suggest Orch OR could explain the mechanism of general anesthesia (acting on microtubules) and address deep questions about consciousness, non-computability, and free will.174 Some propose links between microtubule vibrations and EEG rhythms.174 - Critiques and Challenges: The most significant challenge is decoherence. Critics argue that the brain's "warm, wet, and noisy" environment would cause any quantum superposition within microtubules to collapse almost instantaneously (on femtosecond timescales) due to interactions with the surrounding medium, far too quickly to influence slower neuronal processes.175 While proponents have contested these calculations and proposed shielding mechanisms 175, the decoherence problem remains a major hurdle. Other criticisms include the theory's perceived lack of explanatory power compared to classical neuroscience approaches 175, the speculative nature of the OR mechanism itself 182, and the absence of direct, unambiguous experimental evidence for functional quantum computation occurring within neurons in vivo.175 - Alternative Quantum Models: Theories proposing quantum effects at other levels, such as involving electromagnetic fields or nuclear spins in molecules mediating neurotransmission, also exist but face similar challenges regarding biological plausibility and experimental verification.184 - Current Status and Challenges: Investigating potential sub-millisecond or quantum effects in the brain requires developing novel experimental techniques capable of probing molecular and quantum dynamics at extremely fast timescales within living tissue without causing disruption. Theoretical models must generate specific, falsifiable predictions amenable to experimental testing. At present, the functional relevance of such phenomena remains highly speculative and largely unsupported by direct empirical evidence.175 - Synthesis: While neuroscience has established the importance of precise spike timing at the millisecond scale for information coding 168, the hypothesis that distinct computational mechanisms operating at sub-millisecond timescales play a functional role lacks substantial experimental backing. Similarly, theories invoking quantum mechanics in brain function, such as Orch OR 174, face formidable theoretical obstacles, most notably the problem of maintaining quantum coherence in the biological environment 175, and suffer from a critical lack of direct experimental validation. Proving that such ultra-fast or quantum processes are not just physically possible but are actively utilized by the brain for computation or generating consciousness would require revolutionary experimental evidence far beyond current capabilities. The burden of proof remains exceptionally high for these non-standard hypotheses. B. Bridging Continuous Integration and Discrete Neural Events (RQ8 Analysis) A fundamental puzzle in neuroscience is how the brain transitions from processing seemingly continuous streams of sensory information over time to generating functionally discrete outputs, such as perceptual decisions or motor commands. Is this apparent discreteness an artifact of our experimental paradigms, or does it reflect underlying neural mechanisms like attractor dynamics, thresholding, or oscillatory phase-locking? - The Continuous-Discrete Interface: Natural sensory inputs and behaviors unfold continuously.185 Neural activity itself exhibits dynamics across a range of continuous timescales.185 Yet, many cognitive functions appear discrete: we categorize objects, choose between distinct options, and initiate specific actions at particular moments.188 Understanding the neural mechanisms governing this transformation is crucial. Traditional trial-based experiments often impose discreteness, potentially obscuring the continuous dynamics.185 - Candidate Mechanisms for Discretization: - Evidence Accumulation to Threshold: Models like the Drift-Diffusion Model (DDM) propose that noisy evidence supporting different options is integrated linearly over time. A discrete decision is triggered when the accumulated evidence for one option reaches a predefined threshold or bound.189 This framework successfully explains reaction time distributions and accuracy in many simple decision tasks.190 Neurophysiologically, it is supported by observations of neurons in parietal and frontal cortices exhibiting ramping activity whose slope correlates with evidence strength and which reaches a relatively stereotyped level around the time of decision commitment.189 These models can be implemented in biologically plausible spiking networks.189 - Attractor Dynamics: Recurrent neural networks can exhibit attractor dynamics, where the network state evolves over time and eventually settles into one of several stable states (point attractors or line/plane attractors).193 These stable states can represent discrete choices or categories. Continuous sensory input drives the network state through its state space until it falls into the basin of attraction of one state, effectively making a categorical decision.193 Attractor models naturally implement winner-take-all competition and can account for phenomena like decision stability, changes of mind (if noise pushes the state between basins), and working memory.197 Evidence accumulation in attractor networks is often nonlinear due to feedback and mutual inhibition.193 The geometry of the attractor landscape (e.g., depth of attractor basins) can relate to decision consistency or confidence.197 - Oscillatory Phase-Locking and Gating: Neural oscillations could impose temporal structure, creating discrete windows for information processing or communication.199 The phase of ongoing oscillations, particularly in lower frequency bands like theta or alpha/beta, has been shown to correlate with perceptual detection thresholds and discrimination accuracy, suggesting that perception might be rhythmically modulated or sampled.199 The hierarchical organization of oscillations, where the phase of slower rhythms modulates the amplitude of faster ones (e.g., delta phase modulating theta amplitude, theta phase modulating gamma amplitude), could provide a mechanism for cyclically gating neuronal excitability and controlling information flow between brain regions.202 Entrainment of neural oscillations to rhythmic external stimuli can align these processing windows with relevant events.199 Specific phase relationships (phase coding) can also carry information independently of firing rate.201 - Urgency Gating: This model offers an alternative to pure temporal integration. It proposes that ramping activity reflects the multiplication of a continuously updated estimate of sensory evidence by a growing "urgency" signal that increases over time within a trial.190 This mechanism can also produce decisions when the modulated signal crosses a threshold, but attributes the ramp primarily to urgency rather than accumulated history. It may be particularly relevant under time pressure.190 - Distinguishing Mechanisms Experimentally: Differentiating these potential mechanisms requires careful experimental design and analysis: - Temporal Resolution: High-temporal-resolution recordings (EEG, MEG, intracranial electrophysiology) are essential to capture the fast dynamics associated with oscillations and state transitions.185 - Task Design: Tasks manipulating evidence strength over time 190, varying the number of choices 195, using continuous stimuli and responses 185, or involving rhythmic inputs 199 can help dissociate model predictions. - Model Comparison: Fitting different computational models (DDM, attractor models, urgency models, oscillatory models) to both behavioral data (accuracy, reaction times) and neural data (firing rates, LFP power/phase, population trajectories) allows for quantitative comparison of how well each mechanism explains the observations.190 - Neural Trajectories: Analyzing the paths of neural population activity in state space can reveal signatures of attractor dynamics (convergence to fixed points, movement along manifolds) versus integration (ramping trajectories).195 Recent findings of sequential neural activity during decision tasks challenge models based solely on persistent activity and suggest dynamic transitions between states might be crucial.196 - Integrated Frameworks: Some theoretical frameworks, like active inference, attempt to formally bridge the gap between discrete processes (like policy selection in an MDP) and the continuous dynamics of sensory inference and motor control required to implement those policies.188 - Synthesis: The apparent transition from continuous neural processing to discrete functional outputs like decisions is likely not governed by a single mechanism but emerges from the interplay of various network properties and computational strategies. Evidence accumulation models capture the relationship between evidence strength and decision time 189, while attractor dynamics provide a framework for categorical choice and stability 193, and oscillatory dynamics may impose temporal discretization or gating.199 These mechanisms are not mutually exclusive and may operate in concert or dominate depending on the specific task, brain region, and timescale involved. For instance, attractor networks can implement evidence accumulation.193 Recent observations of sequential dynamics suggest that transitions between states, rather than convergence to a single stable state, might be fundamental in some decision contexts.196 Furthermore, the distinction between continuous neural dynamics and discrete behavioral outputs might be partly artificial; the critical event may be the continuous neural state crossing a functional threshold for action commitment, rather than an inherently discrete neural computation.185 Understanding this interface requires moving towards continuous experimental paradigms and employing computational models that capture these diverse potential mechanisms. - Table 3: Mechanistic Models for Continuous-to-Discrete Processing | | | | | | |---|---|---|---|---| |Mechanism|Core Principle|Key Neural Signature(s)|Supporting Evidence/Snippets|Key Experimental Distinctions| |Evidence Accumulation (DDM)|Linear integration of noisy evidence to a fixed threshold.|Gradually ramping firing rates (slope related to evidence strength); stereotyped activity level at decision time.|Ramping activity in LIP, FEF, etc. 189; fits RT/accuracy data.193|Predicts linear accumulation; test with time-varying evidence 190; analyze single-trial trajectories.| |Attractor Dynamics|Network settles into one of several stable states (attractors) representing choices.|Convergence to distinct population activity states; potentially abrupt transitions; hysteresis; sustained activity (for point attractors).|Models explain categorical choice, working memory, changes of mind 193; some neural data consistent with discrete states.198|Predicts nonlinear accumulation 193; test stability/perturbation responses; analyze state-space trajectories.195| |Oscillatory Phase-Locking/Gating|Phase of ongoing oscillations creates discrete processing windows or gates information flow.|Correlation between oscillatory phase (e.g., theta, alpha, beta) and perceptual accuracy/timing; phase-amplitude coupling.|Phase predicts perceptual accuracy 199; oscillatory hierarchy modulates excitability 202; phase coding exists.201|Test with rhythmic stimuli 199; analyze phase-locking value, phase-RT correlations; manipulate oscillations (e.g., tACS).| |Urgency Gating|Ramping activity arises from multiplying current evidence estimate by a growing urgency signal.|Ramping activity (similar to DDM under constant evidence); potentially distinct response to time-varying evidence.|Explains some behavioral data better than DDM when evidence changes mid-trial.190|Test with time-varying evidence, manipulate time pressure/urgency signals; compare model fits.190| |Sequential Dynamics|Information represented/accumulated across sequences of transiently active neurons.|Choice-specific sequences of neuronal firing; lack of persistent activity in single neurons.|Observed in rodent decision tasks.196|Analyze population dynamics over time; test models with sequence-based computation (e.g., competing chains, bump attractors 196).| V. Synthesis and Future Directions The research frontiers explored in this report collectively challenge simplistic views of neural computation and memory. They highlight a move towards understanding the brain as a system utilizing diverse information carriers beyond spikes, operating with dynamic and context-sensitive codes, storing memories through multi-level plasticity mechanisms, and employing sophisticated strategies for complex behaviors and fundamental computations. A key theme emerging across these diverse topics is the inadequacy of focusing on single modalities or mechanisms in isolation. Understanding neural encoding requires considering the interplay of spikes, LFPs, glial activity, and field effects (RQ1), all modulated by network context (RQ2). Memory persistence likely involves a combination of synaptic, intracellular, epigenetic, and potentially non-synaptic mechanisms (RQ3, RQ4), whose interactions shape the reconstructive nature and subjective experience of recall (RQ5). Complex behaviors like ant navigation demonstrate the power of integrating multiple sensory modalities within a flexible, learned hierarchy (RQ6). Even fundamental questions about processing speed and the continuous-discrete interface point towards multiple interacting mechanisms rather than single solutions (RQ7, RQ8). This interconnectedness underscores the need for integrated experimental and theoretical approaches. Progress hinges on developing and deploying multi-modal technologies capable of simultaneously recording diverse signal types (electrical, chemical, optical) across multiple spatial and temporal scales.1 Concurrently, computational models must evolve beyond simplified assumptions to incorporate biological realism, context dependency, generative processes, and the dynamics of multiple interacting mechanisms.8 Significant challenges remain. Technological limitations constrain our ability to probe intracellular molecular dynamics or potential quantum effects in vivo non-invasively. Modeling the sheer complexity of context-dependent network dynamics is computationally demanding. Bridging the vast gap between molecular mechanisms, network activity, and subjective cognitive experience (like memory qualia) remains a central hurdle. Future research directions stemming from the reviewed questions include: - Developing novel sensors and probes: Creating tools for high-resolution, chronic, multi-modal recording, including reliable in vivo sensing of neuromodulators, glial activity, and potentially subtle electrical field effects. - Refining computational models: Building models that explicitly incorporate network context (neuromodulation, oscillations, glial state, behavioral state), learn generative representations, and bridge different levels of analysis (from biophysical detail to cognitive function). Testing these models against diverse, naturalistic datasets is crucial. - Targeting non-synaptic mechanisms: Designing experiments using tools like CRISPR 85 and targeted intracellular reporters/manipulators to specifically investigate the role of epigenetic modifications, intrinsic excitability changes, and other non-synaptic factors in long-term memory storage and persistence, while carefully controlling for synaptic effects. - Causal validation: Employing causal perturbation techniques (optogenetics, chemogenetics, TMS, focused ultrasound) guided by computational model predictions to test the functional necessity of specific signals (e.g., LFP bands, glial signals), network states, or proposed mechanisms (e.g., attractor dynamics vs. integration) for specific cognitive functions or behaviors.120 - Embracing continuous paradigms: Moving beyond traditional trial-based experiments towards more naturalistic, continuous tasks to better capture the ongoing dynamics of brain activity and behavior.185 The questions explored here represent the cutting edge of neuroscience, demanding a departure from established paradigms and embracing complexity. While definitive answers remain elusive for many of these RQs, the skeptical, inquiry-driven approach, fueled by technological innovation and theoretical advances, continues to push the frontiers of our understanding of the brain. The pursuit of these questions promises not only deeper insights into neural computation, memory, and behavior but also potential avenues for addressing neurological and psychiatric disorders characterized by disruptions in these fundamental processes. #### Works cited 1. (PDF) Advanced Neural Probe Sensors toward Multi‐Modal Sensing ..., accessed April 20, 2025, [https://www.researchgate.net/publication/387124637_Advanced_Neural_Probe_Sensors_toward_Multi-Modal_Sensing_and_Modulation_Design_Integration_and_Applications](https://www.researchgate.net/publication/387124637_Advanced_Neural_Probe_Sensors_toward_Multi-Modal_Sensing_and_Modulation_Design_Integration_and_Applications) 2. From Neurons to Cognition: Technologies for Precise Recording of Neural Activity Underlying Behavior - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10521756/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10521756/) 3. Multimodal integrated flexible neural probe for in situ monitoring of EEG and lactic acid, accessed April 20, 2025, [https://pubs.rsc.org/en/content/articlehtml/2024/ra/d4ra06336h](https://pubs.rsc.org/en/content/articlehtml/2024/ra/d4ra06336h) 4. Multimodal in vivo recording using transparent graphene microelectrodes illuminates spatiotemporal seizure dynamics at the microscale, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7846732/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7846732/) 5. Multifunctional and Flexible Neural Probe with Thermally Drawn Fibers for Bidirectional Synaptic Probing in the Brain | ACS Nano - ACS Publications, accessed April 20, 2025, [https://pubs.acs.org/doi/10.1021/acsnano.4c02578](https://pubs.acs.org/doi/10.1021/acsnano.4c02578) 6. Recent developments in multifunctional neural probes for simultaneous neural recording and modulation - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/366850305_Recent_developments_in_multifunctional_neural_probes_for_simultaneous_neural_recording_and_modulation](https://www.researchgate.net/publication/366850305_Recent_developments_in_multifunctional_neural_probes_for_simultaneous_neural_recording_and_modulation) 7. Novel electrode technologies for neural recordings - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6531316/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6531316/) 8. [2503.16567] Exploring Deep Learning Models for EEG Neural Decoding - arXiv, accessed April 20, 2025, [https://arxiv.org/abs/2503.16567](https://arxiv.org/abs/2503.16567) 9. Neural decoding reveals specialized kinematic tuning after an ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10591925/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10591925/) 10. Computational Models to Study Language Processing in the Human Brain - arXiv, accessed April 20, 2025, [https://arxiv.org/pdf/2403.13368](https://arxiv.org/pdf/2403.13368) 11. Computational Models to Study Language Processing in the Human Brain: A Survey - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2403.13368v1](https://arxiv.org/html/2403.13368v1) 12. Navigating Brain Language Representations: A Comparative Analysis of Neural Language Models and Psychologically Plausible Models - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2404.19364](https://arxiv.org/html/2404.19364) 13. Local field potentials reflect cortical population dynamics in a region ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC9470163/](https://pmc.ncbi.nlm.nih.gov/articles/PMC9470163/) 14. Relationships between spike-free local field potentials and spike ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC3331669/](https://pmc.ncbi.nlm.nih.gov/articles/PMC3331669/) 15. Adaptive Spike-Artifact Removal from Local Field Potentials Uncovers Prominent Beta and Gamma Band Neuronal Synchronization - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7242114/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7242114/) 16. Reliability of directional information in unsorted spikes and local ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4142142/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4142142/) 17. Recent Advances in Brain Physiology and Cognitive Processing ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC3115287/](https://pmc.ncbi.nlm.nih.gov/articles/PMC3115287/) 18. Glial cells in neuronal network function - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC2894949/](https://pmc.ncbi.nlm.nih.gov/articles/PMC2894949/) 19. www.mdpi.com, accessed April 20, 2025, [https://www.mdpi.com/1422-0067/25/3/1899#:~:text=Astrocytes%20interact%20with%20synaptic%20neuronal,regulating%20synaptic%20transmission%20and%20plasticity.](https://www.mdpi.com/1422-0067/25/3/1899#:~:text=Astrocytes%20interact%20with%20synaptic%20neuronal,regulating%20synaptic%20transmission%20and%20plasticity.) 20. The Role of Astrocytes in the Neurorepair Process - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/cell-and-developmental-biology/articles/10.3389/fcell.2021.665795/full](https://www.frontiersin.org/journals/cell-and-developmental-biology/articles/10.3389/fcell.2021.665795/full) 21. Astrocytic modulation of neuronal signalling - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/network-physiology/articles/10.3389/fnetp.2023.1205544/full](https://www.frontiersin.org/journals/network-physiology/articles/10.3389/fnetp.2023.1205544/full) 22. Emerging roles of astrocytes in neural circuit development - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4431630/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4431630/) 23. Astrocyte Calcium Waves: What They Are and What They Do - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC2605018/](https://pmc.ncbi.nlm.nih.gov/articles/PMC2605018/) 24. Calcium signalling in glial cells - PubMed, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/8843604/](https://pubmed.ncbi.nlm.nih.gov/8843604/) 25. Glial calcium - PubMed, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/8244537/](https://pubmed.ncbi.nlm.nih.gov/8244537/) 26. Neuron-glia networks: integral gear of brain function - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4222327/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4222327/) 27. Essential Role of Astrocytes in Learning and Memory - MDPI, accessed April 20, 2025, [https://www.mdpi.com/1422-0067/25/3/1899](https://www.mdpi.com/1422-0067/25/3/1899) 28. Properties of Synaptically Evoked Astrocyte Calcium Signal Reveal Synaptic Information Processing by Astrocytes | Journal of Neuroscience, accessed April 20, 2025, [https://www.jneurosci.org/content/25/9/2192](https://www.jneurosci.org/content/25/9/2192) 29. Complementary encoding of spatial information in hippocampal astrocytes - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC8893713/](https://pmc.ncbi.nlm.nih.gov/articles/PMC8893713/) 30. Revisiting astrocytic calcium signaling in the brain - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC11670731/](https://pmc.ncbi.nlm.nih.gov/articles/PMC11670731/) 31. Calcium Signals in Astrocyte Microdomains, a Decade of Great Advances - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC8216559/](https://pmc.ncbi.nlm.nih.gov/articles/PMC8216559/) 32. Neuronal calcium signaling: function and dysfunction - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC11113927/](https://pmc.ncbi.nlm.nih.gov/articles/PMC11113927/) 33. Decoding Calcium Signaling Across the Nucleus - American Journal of Physiology, accessed April 20, 2025, [https://journals.physiology.org/doi/full/10.1152/physiol.00056.2013](https://journals.physiology.org/doi/full/10.1152/physiol.00056.2013) 34. Annihilation of action potentials induces electrical coupling between neurons | eLife, accessed April 20, 2025, [https://elifesciences.org/articles/88335](https://elifesciences.org/articles/88335) 35. Realistic modeling of mesoscopic ephaptic coupling in the human ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7289436/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7289436/) 36. A mathematical model of ephaptic interactions in neuronal fiber pathways: Could there be more than transmission along the tracts? - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7462434/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7462434/) 37. In vivo ephaptic coupling allows memory network formation ..., accessed April 20, 2025, [https://academic.oup.com/cercor/article/33/17/9877/7220593](https://academic.oup.com/cercor/article/33/17/9877/7220593) 38. Ephaptic coupling - Wikipedia, accessed April 20, 2025, [https://en.wikipedia.org/wiki/Ephaptic_coupling](https://en.wikipedia.org/wiki/Ephaptic_coupling) 39. Ephaptic Coupling Promotes Synchronous Firing of Cerebellar Purkinje Cells - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7513896/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7513896/) 40. Ephaptic coupling in white matter fibre bundles modulates axonal transmission delays, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7895385/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7895385/) 41. Annihilation of action potentials induces electrical coupling between neurons - eLife, accessed April 20, 2025, [https://elifesciences.org/reviewed-preprints/88335v3](https://elifesciences.org/reviewed-preprints/88335v3) 42. Annihilation of action potentials induces electrical coupling between neurons - eLife, accessed April 20, 2025, [https://elifesciences.org/reviewed-preprints/88335v2](https://elifesciences.org/reviewed-preprints/88335v2) 43. Multi-region calcium imaging in freely behaving mice with ultra-compact head-mounted fluorescence microscopes - Oxford Academic, accessed April 20, 2025, [https://academic.oup.com/nsr/article/11/1/nwad294/7438875](https://academic.oup.com/nsr/article/11/1/nwad294/7438875) 44. Stable sound decoding despite modulated sound representation in ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10515783/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10515783/) 45. Context-dependent neural preparation for information relevance vs ..., accessed April 20, 2025, [https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00302/124456/Context-dependent-neural-preparation-for](https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00302/124456/Context-dependent-neural-preparation-for) 46. The potential of convolutional neural networks for identifying neural states based on electrophysiological signals: experiments on synthetic and real patient data - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10272439/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10272439/) 47. Effects of Noise Correlations on Information Encoding and Decoding ..., accessed April 20, 2025, [https://journals.physiology.org/doi/full/10.1152/jn.00919.2005](https://journals.physiology.org/doi/full/10.1152/jn.00919.2005) 48. Information-Limiting Correlations in Large Neural Populations - Journal of Neuroscience, accessed April 20, 2025, [https://www.jneurosci.org/content/40/8/1668](https://www.jneurosci.org/content/40/8/1668) 49. Information-Limiting Correlations in Neural Populations: The Devil Is in the Details - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7548692/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7548692/) 50. Strong, but not weak, noise correlations are beneficial for population coding - bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2024.06.26.600826v1.full-text](https://www.biorxiv.org/content/10.1101/2024.06.26.600826v1.full-text) 51. Effects of noise correlations on encoding and decoding. A: components... | Download Scientific Diagram - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/figure/Effects-of-noise-correlations-on-encoding-and-decoding-A-components-of-d-2-vs-the_fig2_7223783](https://www.researchgate.net/figure/Effects-of-noise-correlations-on-encoding-and-decoding-A-components-of-d-2-vs-the_fig2_7223783) 52. Modeling the diverse effects of divisive normalization on noise correlations - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10715670/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10715670/) 53. Effects of Noise Correlations on Information Encoding and Decoding | Journal of Neurophysiology | American Physiological Society, accessed April 20, 2025, [https://journals.physiology.org/doi/prev/20171106-aop/abs/10.1152/jn.00919.2005](https://journals.physiology.org/doi/prev/20171106-aop/abs/10.1152/jn.00919.2005) 54. When and Why Noise Correlations Are Important in Neural Decoding - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6618421/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6618421/) 55. A latent variable approach to decoding neural population activity - bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2020.01.06.896423v1.full-text](https://www.biorxiv.org/content/10.1101/2020.01.06.896423v1.full-text) 56. Neural models for detection and classification of brain states and transitions - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC11986132/](https://pmc.ncbi.nlm.nih.gov/articles/PMC11986132/) 57. Neuromodulation of Neural Oscillations in Health and Disease, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/36979063/](https://pubmed.ncbi.nlm.nih.gov/36979063/) 58. Brain-to-Text Decoding with Context-Aware Neural Representations and Large Language Models - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2411.10657v1](https://arxiv.org/html/2411.10657v1) 59. Brain-to-Text Benchmark '24: Lessons Learned - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2412.17227v1](https://arxiv.org/html/2412.17227v1) 60. Brain Foundation Models: A Survey on Advancements in Neural Signal Processing and Brain Discovery - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2503.00580v1](https://arxiv.org/html/2503.00580v1) 61. Brain-inspired Computing Based on Deep Learning for Human-computer Interaction: A Review - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2312.07213v4/](https://arxiv.org/html/2312.07213v4/) 62. Invariant neural subspaces maintained by feedback modulation - eLife, accessed April 20, 2025, [https://elifesciences.org/articles/76096.pdf](https://elifesciences.org/articles/76096.pdf) 63. Deep learning approaches for neural decoding across architectures and recording modalities - Oxford Academic, accessed April 20, 2025, [https://academic.oup.com/bib/article/22/2/1577/6054827](https://academic.oup.com/bib/article/22/2/1577/6054827) 64. The computational and neural bases of context-dependent learning - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10348919/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10348919/) 65. Testing for context-dependent changes in neural encoding in naturalistic experiments - arXiv, accessed April 20, 2025, [https://arxiv.org/abs/2211.09295](https://arxiv.org/abs/2211.09295) 66. Associative Memories via Predictive Coding - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7612799/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7612799/) 67. (PDF) The Generative Adversarial Brain - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/335887590_The_Generative_Adversarial_Brain](https://www.researchgate.net/publication/335887590_The_Generative_Adversarial_Brain) 68. Memory engrams: Recalling the past and imagining the future - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7577560/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7577560/) 69. Understanding the physical basis of memory: Molecular mechanisms of the engram - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC9065729/](https://pmc.ncbi.nlm.nih.gov/articles/PMC9065729/) 70. Tissue clearing applications in memory engram research - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10493294/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10493294/) 71. Exploring the memory: existing activity-dependent tools to tag and manipulate engram cells, accessed April 20, 2025, [https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2023.1279032/full](https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2023.1279032/full) 72. Exploring the memory: existing activity-dependent tools to tag and manipulate engram cells - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10800721/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10800721/) 73. Exploring the memory: existing activity-dependent tools to tag and manipulate engram cells, accessed April 20, 2025, [https://ouci.dntb.gov.ua/en/works/4wBjjWY7/](https://ouci.dntb.gov.ua/en/works/4wBjjWY7/) 74. Optogenetic stimulation of a hippocampal engram activates fear memory recall, accessed April 20, 2025, [https://www.researchgate.net/publication/221960129_Optogenetic_stimulation_of_a_hippocampal_engram_activates_fear_memory_recall](https://www.researchgate.net/publication/221960129_Optogenetic_stimulation_of_a_hippocampal_engram_activates_fear_memory_recall) 75. Identification and optogenetic manipulation of memory engrams in the hippocampus, accessed April 20, 2025, [https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2013.00226/full](https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2013.00226/full) 76. Engram-specific transcriptome profiling of contextual memory ..., accessed April 20, 2025, [https://repub.eur.nl/pub/116902/Repub_116902_O-A.pdf](https://repub.eur.nl/pub/116902/Repub_116902_O-A.pdf) 77. Selective Increase of Correlated Activity in Arc-Positive Neurons after Chemically Induced Long-Term Potentiation in Cultured Hippocampal Neurons | eNeuro, accessed April 20, 2025, [https://www.eneuro.org/content/8/6/ENEURO.0540-20.2021](https://www.eneuro.org/content/8/6/ENEURO.0540-20.2021) 78. What is memory? The present state of the engram - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4874022/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4874022/) 79. Manipulating a “Cocaine Engram” in Mice | Journal of Neuroscience, accessed April 20, 2025, [https://www.jneurosci.org/content/34/42/14115](https://www.jneurosci.org/content/34/42/14115) 80. Reconsolidation and the Dynamic Nature of Memory - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4588064/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4588064/) 81. On the Involvement of BDNF Signaling in Memory Reconsolidation - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2019.00383/full](https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2019.00383/full) 82. Retrieval-Dependent Mechanisms Affecting Emotional Memory Persistence: Reconsolidation, Extinction, and the Space in Between - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2020.574358/full](https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2020.574358/full) 83. A meta-analysis of the effect of protein synthesis inhibitors on rodent fear conditioning, accessed April 20, 2025, [https://peercommunityjournal.org/articles/10.24072/pcjournal.339/](https://peercommunityjournal.org/articles/10.24072/pcjournal.339/) 84. Remote memory engrams are controlled by encoding-specific tau ..., accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2023.12.01.569663v1.full](https://www.biorxiv.org/content/10.1101/2023.12.01.569663v1.full) 85. Conditional Genome Editing in the Mammalian Brain Using CRISPR ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7954951/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7954951/) 86. Neurochemical mechanisms for memory processing during sleep ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6879745/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6879745/) 87. Homeostatic plasticity and excitation-inhibition balance: The good, the bad and the ugly - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC9477500/](https://pmc.ncbi.nlm.nih.gov/articles/PMC9477500/) 88. Homeostatic plasticity - Wikipedia, accessed April 20, 2025, [https://en.wikipedia.org/wiki/Homeostatic_plasticity](https://en.wikipedia.org/wiki/Homeostatic_plasticity) 89. Editorial: Homeostatic Synaptic Plasticity: From Synaptic Circuit Assembly to Neurological Disorders - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2021.695313/full](https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2021.695313/full) 90. Mechanisms of Homeostatic Synaptic Plasticity in vivo - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2019.00520/full](https://www.frontiersin.org/journals/cellular-neuroscience/articles/10.3389/fncel.2019.00520/full) 91. Sleep, Plasticity and Memory from Molecules to Whole-Brain Networks - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4263505/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4263505/) 92. System consolidation of memory during sleep - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC3278619/](https://pmc.ncbi.nlm.nih.gov/articles/PMC3278619/) 93. Can sleep protect memories from catastrophic forgetting? - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7440920/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7440920/) 94. Nonsynaptic plasticity - Wikipedia, accessed April 20, 2025, [https://en.wikipedia.org/wiki/Nonsynaptic_plasticity](https://en.wikipedia.org/wiki/Nonsynaptic_plasticity) 95. Memory engram stability and flexibility - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC11525749/](https://pmc.ncbi.nlm.nih.gov/articles/PMC11525749/) 96. The Future of Memory: Remembering, Imagining, and the Brain - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC3815616/](https://pmc.ncbi.nlm.nih.gov/articles/PMC3815616/) 97. Interplay of hippocampus and prefrontal cortex in memory - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC3789138/](https://pmc.ncbi.nlm.nih.gov/articles/PMC3789138/) 98. Evidence supporting a time-limited hippocampal role in retrieving autobiographical memories | PNAS, accessed April 20, 2025, [https://www.pnas.org/doi/10.1073/pnas.2023069118](https://www.pnas.org/doi/10.1073/pnas.2023069118) 99. The cognitive neuroscience of constructive memory: remembering ..., accessed April 20, 2025, [https://royalsocietypublishing.org/doi/abs/10.1098/rstb.2007.2087](https://royalsocietypublishing.org/doi/abs/10.1098/rstb.2007.2087) 100. Effects of Internal and External Vividness on Hippocampal ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4938785/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4938785/) 101. Hippocampal subfields and their neocortical interactions during ..., accessed April 20, 2025, [https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00105/119707/Hippocampal-subfields-and-their-neocortical](https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00105/119707/Hippocampal-subfields-and-their-neocortical) 102. Editorial: Multimodal approaches to investigating neural dynamics in cognition and related clinical conditions: integrating EEG, MEG, and fMRI data - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/systems-neuroscience/articles/10.3389/fnsys.2025.1495018/full](https://www.frontiersin.org/journals/systems-neuroscience/articles/10.3389/fnsys.2025.1495018/full) 103. Magnetoencephalography in the study of brain dynamics - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4370437/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4370437/) 104. Comparison between EEG and MEG of static and dynamic resting-state networks - PubMed, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/39230193/](https://pubmed.ncbi.nlm.nih.gov/39230193/) 105. Human memory retention and recall processes. A review of EEG and fMRI studies - PubMed, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/24141456/](https://pubmed.ncbi.nlm.nih.gov/24141456/) 106. Predictive Processing: A Circuit Approach to Psychosis - Annual Reviews, accessed April 20, 2025, [https://www.annualreviews.org/content/journals/10.1146/annurev-neuro-100223-121214](https://www.annualreviews.org/content/journals/10.1146/annurev-neuro-100223-121214) 107. Theory of Mind: A Neural Prediction Problem - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4041537/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4041537/) 108. Evaluating the neurophysiological evidence for predictive processing as a model of perception - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7187369/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7187369/) 109. Evidence and implications of abnormal predictive coding in dementia - Oxford Academic, accessed April 20, 2025, [https://academic.oup.com/brain/article/144/11/3311/6317665](https://academic.oup.com/brain/article/144/11/3311/6317665) 110. Dynamic Predictive Coding: A Model of Hierarchical Sequence Learning and Prediction in the Neocortex | bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2022.06.23.497415v4.full](https://www.biorxiv.org/content/10.1101/2022.06.23.497415v4.full) 111. Illusory Motion Reproduced by Deep Neural Networks Trained for Prediction - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2018.00345/full](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2018.00345/full) 112. Prediction error and repetition suppression have distinct effects on neural representations of visual information | eLife, accessed April 20, 2025, [https://elifesciences.org/articles/33123](https://elifesciences.org/articles/33123) 113. [2109.08063] Associative Memories via Predictive Coding - arXiv, accessed April 20, 2025, [https://arxiv.org/abs/2109.08063](https://arxiv.org/abs/2109.08063) 114. A neurophenomenological model for the role of the hippocampus in temporal consciousness. Evidence from confabulation - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4549641/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4549641/) 115. Models of confabulation: A critical review and new framework - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/5437163_Models_of_confabulation_A_critical_review_and_new_framework](https://www.researchgate.net/publication/5437163_Models_of_confabulation_A_critical_review_and_new_framework) 116. The cognitive neuroscience of confabulation: A review and a model - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/295608835_The_cognitive_neuroscience_of_confabulation_A_review_and_a_model](https://www.researchgate.net/publication/295608835_The_cognitive_neuroscience_of_confabulation_A_review_and_a_model) 117. Confabulation Theory - National Academic Digital Library of Ethiopia, accessed April 20, 2025, [http://ndl.ethernet.edu.et/bitstream/123456789/49680/1/214.pdf](http://ndl.ethernet.edu.et/bitstream/123456789/49680/1/214.pdf) 118. Confabulation theory (computational intelligence) - Scholarpedia, accessed April 20, 2025, [http://www.scholarpedia.org/article/Confabulation_theory_(computational_intelligence)](http://www.scholarpedia.org/article/Confabulation_theory_\(computational_intelligence\)) 119. Hallucination or Confabulation? Neuroanatomy as metaphor in Large Language Models, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10619792/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10619792/) 120. Model of light-sensitive channels. (A) The dynamics of the three ..., accessed April 20, 2025, [https://www.researchgate.net/figure/Model-of-light-sensitive-channels-A-The-dynamics-of-the-three-state-model-for-ChR2_fig1_342406157](https://www.researchgate.net/figure/Model-of-light-sensitive-channels-A-The-dynamics-of-the-three-state-model-for-ChR2_fig1_342406157) 121. TMS for Memory Loss | Memory Loss Treatment - NeuroSpa Brain Rejuvenation, accessed April 20, 2025, [https://neurospabrain.com/memory-loss/](https://neurospabrain.com/memory-loss/) 122. Testing network properties of episodic memory using non-invasive ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7138212/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7138212/) 123. The physiological effects of noninvasive brain stimulation fundamentally differ across the human cortex - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6994208/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6994208/) 124. Brain stimulation boosts memory replay, accuracy - Northwestern Now, accessed April 20, 2025, [https://news.northwestern.edu/stories/2021/02/brain-stimulation-boosts-memory-replay-accuracy/](https://news.northwestern.edu/stories/2021/02/brain-stimulation-boosts-memory-replay-accuracy/) 125. Transcranial Magnetic Stimulation Trains at 1 Hz ... - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2021.696793/full](https://www.frontiersin.org/journals/human-neuroscience/articles/10.3389/fnhum.2021.696793/full) 126. Review of Noninvasive or Minimally Invasive Deep Brain Stimulation - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2021.820017/full](https://www.frontiersin.org/journals/behavioral-neuroscience/articles/10.3389/fnbeh.2021.820017/full) 127. Theta power and theta-gamma coupling support spatial memory retrieval - bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/732735v1](https://www.biorxiv.org/content/10.1101/732735v1) 128. A comparative study of navigation behaviours in ants | bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2024.10.24.619962v1.full-text](https://www.biorxiv.org/content/10.1101/2024.10.24.619962v1.full-text) 129. Insect Navigation: Neural Basis to Behavior | Oxford Research Encyclopedia of Neuroscience, accessed April 20, 2025, [https://oxfordre.com/neuroscience/display/10.1093/acrefore/9780190264086.001.0001/acrefore-9780190264086-e-175?p=emailAUv0ljDXgJPZk&d=/10.1093/acrefore/9780190264086.001.0001/acrefore-9780190264086-e-175](https://oxfordre.com/neuroscience/display/10.1093/acrefore/9780190264086.001.0001/acrefore-9780190264086-e-175?p=emailAUv0ljDXgJPZk&d=/10.1093/acrefore/9780190264086.001.0001/acrefore-9780190264086-e-175) 130. Australian Journal of Zoology - CSIRO PUBLISHING, accessed April 20, 2025, [https://www.publish.csiro.au/zo/ZO16046](https://www.publish.csiro.au/zo/ZO16046) 131. How to Navigate in Different Environments and Situations ... - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2018.00841/full](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2018.00841/full) 132. The Sensory Ecology of Ant Navigation: From Natural Environments to Neural Mechanisms - Annual Reviews, accessed April 20, 2025, [https://www.annualreviews.org/doi/pdf/10.1146/annurev-ento-010715-023703](https://www.annualreviews.org/doi/pdf/10.1146/annurev-ento-010715-023703) 133. myrmecologicalnews.org, accessed April 20, 2025, [https://myrmecologicalnews.org/cms/index.php?option=com_download&view=download&filename=volume29/mn29_1-20_printable.pdf&format=raw](https://myrmecologicalnews.org/cms/index.php?option=com_download&view=download&filename=volume29/mn29_1-20_printable.pdf&format=raw) 134. Learning walks in an Australian desert ant, Melophorus bagoti - Company of Biologists journals, accessed April 20, 2025, [https://journals.biologists.com/jeb/article-pdf/doi/10.1242/jeb.242177/2090159/jeb242177.pdf](https://journals.biologists.com/jeb/article-pdf/doi/10.1242/jeb.242177/2090159/jeb242177.pdf) 135. The architecture of the desert ant's navigational toolkit (Hymenoptera: Formicidae) - Rüdiger WEHNER - AntWiki, accessed April 20, 2025, [https://www.antwiki.org/wiki/images/9/97/Wehner%2C_R._2009._The_architecture_of_the_desert_ant%27s_navigational_toolkit.pdf](https://www.antwiki.org/wiki/images/9/97/Wehner%2C_R._2009._The_architecture_of_the_desert_ant%27s_navigational_toolkit.pdf) 136. Experimental ethology of learning in desert ants: Becoming expert navigators - PubMed, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/30529645/](https://pubmed.ncbi.nlm.nih.gov/30529645/) 137. Multimodal influences on learning walks in desert ants (Cataglyphis ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7392947/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7392947/) 138. Optimal multiguidance integration in insect navigation | PNAS, accessed April 20, 2025, [https://www.pnas.org/doi/10.1073/pnas.1721668115](https://www.pnas.org/doi/10.1073/pnas.1721668115) 139. Principles of insect path integration - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6462409/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6462409/) 140. Path integration as the basic navigation mechanism of the desert ant Cataglyphis fortis (FOREL, 1902) (Hymenoptera - Myrmecological News, accessed April 20, 2025, [https://myrmecologicalnews.org/cms/index.php?option=com_download&view=download&filename=volume11/mn11_53-62_printable.pdf&format=raw](https://myrmecologicalnews.org/cms/index.php?option=com_download&view=download&filename=volume11/mn11_53-62_printable.pdf&format=raw) 141. (PDF) Principles of Insect Path Integration - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/327566710_Principles_of_Insect_Path_Integration](https://www.researchgate.net/publication/327566710_Principles_of_Insect_Path_Integration) 142. Ant visual route navigation: How the fine details of behaviour ..., accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2025.01.16.633342v1.full-text](https://www.biorxiv.org/content/10.1101/2025.01.16.633342v1.full-text) 143. Compensation to visual impairments and behavioral plasticity in ..., accessed April 20, 2025, [https://www.pnas.org/doi/10.1073/pnas.2410908121](https://www.pnas.org/doi/10.1073/pnas.2410908121) 144. A Model of Ant Route Navigation Driven by Scene Familiarity | PLOS Computational Biology, accessed April 20, 2025, [https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1002336](https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1002336) 145. Similarities and differences in path integration and search in two species of desert ants inhabiting a visually rich and a visually barren habitat - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/303290852_Similarities_and_differences_in_path_integration_and_search_in_two_species_of_desert_ants_inhabiting_a_visually_rich_and_a_visually_barren_habitat](https://www.researchgate.net/publication/303290852_Similarities_and_differences_in_path_integration_and_search_in_two_species_of_desert_ants_inhabiting_a_visually_rich_and_a_visually_barren_habitat) 146. Desert Ants Learn Vibration and Magnetic Landmarks - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC3296771/](https://pmc.ncbi.nlm.nih.gov/articles/PMC3296771/) 147. Importance of magnetic information for neuronal plasticity in desert ..., accessed April 20, 2025, [https://www.pnas.org/doi/10.1073/pnas.2320764121](https://www.pnas.org/doi/10.1073/pnas.2320764121) 148. Internal Compass Helps Ants Navigate Long Distances - Tellus - USDA, accessed April 20, 2025, [https://tellus.ars.usda.gov/stories/articles/internal-compass-helps-ants-navigate-long-distances](https://tellus.ars.usda.gov/stories/articles/internal-compass-helps-ants-navigate-long-distances) 149. Desert Ants Learn Vibration and Magnetic Landmarks - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/221698753_Desert_Ants_Learn_Vibration_and_Magnetic_Landmarks](https://www.researchgate.net/publication/221698753_Desert_Ants_Learn_Vibration_and_Magnetic_Landmarks) 150. Ants Can Learn to Navigate with Chill Magnetic Vibes - VICE, accessed April 20, 2025, [https://www.vice.com/en/article/ants-can-learn-to-navigate-with-chill-magnetic-vibes/](https://www.vice.com/en/article/ants-can-learn-to-navigate-with-chill-magnetic-vibes/) 151. EU research shows magnetic fields, vibrations help ants navigate - CORDIS, accessed April 20, 2025, [https://cordis.europa.eu/article/id/34395-eu-research-shows-magnetic-fields-vibrations-help-ants-navigate](https://cordis.europa.eu/article/id/34395-eu-research-shows-magnetic-fields-vibrations-help-ants-navigate) 152. Wind and Sky as Compass Cues in Desert Ant Navigation | Request PDF - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/6446629_Wind_and_Sky_as_Compass_Cues_in_Desert_Ant_Navigation](https://www.researchgate.net/publication/6446629_Wind_and_Sky_as_Compass_Cues_in_Desert_Ant_Navigation) 153. Dynamic multimodal interactions in navigating wood ants: what do ..., accessed April 20, 2025, [https://journals.biologists.com/jeb/article/223/7/jeb221036/223759/Dynamic-multimodal-interactions-in-navigating-wood](https://journals.biologists.com/jeb/article/223/7/jeb221036/223759/Dynamic-multimodal-interactions-in-navigating-wood) 154. Dynamic multimodal interactions in navigating wood ants: What do path details tell us about cue integration? | bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2019.12.31.891689.full](https://www.biorxiv.org/content/10.1101/2019.12.31.891689.full) 155. Multimodal interactions in insect navigation - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7700066/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7700066/) 156. Path Integration Controls Nest-Plume Following in Desert Ants - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/221691879_Path_Integration_Controls_Nest-Plume_Following_in_Desert_Ants](https://www.researchgate.net/publication/221691879_Path_Integration_Controls_Nest-Plume_Following_in_Desert_Ants) 157. International Society for Neuroethology - Bluesky, accessed April 20, 2025, [https://bsky.app/profile/neuroethology.org/post/3lidnxfyuvc2n](https://bsky.app/profile/neuroethology.org/post/3lidnxfyuvc2n) 158. Multisensory integration supports configural learning of a home refuge in the whip spider Phrynus marginemaculatus - Company of Biologists journals, accessed April 20, 2025, [https://journals.biologists.com/jeb/article/224/3/jeb238444/223406/Multisensory-integration-supports-configural](https://journals.biologists.com/jeb/article/224/3/jeb238444/223406/Multisensory-integration-supports-configural) 159. Examples of multimodal interactions in navigating ants. a If an ant... - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/figure/Examples-of-multimodal-interactions-in-navigating-ants-a-If-an-ant-runs-off-her-entire_fig3_340848527](https://www.researchgate.net/figure/Examples-of-multimodal-interactions-in-navigating-ants-a-If-an-ant-runs-off-her-entire_fig3_340848527) 160. Multi-modal cue integration in the black garden ant - Unipd, accessed April 20, 2025, [https://www.research.unipd.it/bitstream/11577/3358900/2/AnimCog_2020.pdf](https://www.research.unipd.it/bitstream/11577/3358900/2/AnimCog_2020.pdf) 161. INSECT NAVIGATION AND CUE PERCEPTION IN ... - ScholarWorks, accessed April 20, 2025, [https://scholarworks.calstate.edu/downloads/n009w827f](https://scholarworks.calstate.edu/downloads/n009w827f) 162. The neuroethology of ant navigation - Figshare, accessed April 20, 2025, [https://figshare.com/ndownloader/files/53580467/1](https://figshare.com/ndownloader/files/53580467/1) 163. Rüdiger Wehner - Publications - Neurotree, accessed April 20, 2025, [https://neurotree.org/neurotree/publications.php?pid=3409](https://neurotree.org/neurotree/publications.php?pid=3409) 164. Optimal multiguidance integration in insect navigation - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC5856556/](https://pmc.ncbi.nlm.nih.gov/articles/PMC5856556/) 165. The role of learning-walk related multisensory experience in rewiring visual circuits in the desert ant brain - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC10354187/](https://pmc.ncbi.nlm.nih.gov/articles/PMC10354187/) 166. Hard limits to cognitive flexibility: ants can learn to ignore but not ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC8214833/](https://pmc.ncbi.nlm.nih.gov/articles/PMC8214833/) 167. Ants prefer the option they are trained to first | Journal of Experimental Biology, accessed April 20, 2025, [https://journals.biologists.com/jeb/article/225/24/jeb243984/286063/Ants-prefer-the-option-they-are-trained-to-first](https://journals.biologists.com/jeb/article/225/24/jeb243984/286063/Ants-prefer-the-option-they-are-trained-to-first) 168. Neural Coding of Natural Stimuli: Information at Sub-Millisecond ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC2265477/](https://pmc.ncbi.nlm.nih.gov/articles/PMC2265477/) 169. Sub-millisecond firing synchrony of closely neighboring ... - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/neural-circuits/articles/10.3389/neuro.04.009.2009/full](https://www.frontiersin.org/journals/neural-circuits/articles/10.3389/neuro.04.009.2009/full) 170. Sub- and Supra-Second Timing in Auditory Perception: Evidence for Cross-Domain Relationships - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2021.812533/full](https://www.frontiersin.org/journals/neuroscience/articles/10.3389/fnins.2021.812533/full) 171. Testing the state-dependent model of subsecond time perception against experimental evidence - eLife, accessed April 20, 2025, [https://elifesciences.org/reviewed-preprints/94418](https://elifesciences.org/reviewed-preprints/94418) 172. Neural networks engaged in milliseconds and seconds time processing: evidence from transcranial magnetic stimulation and patients with cortical or subcortical dysfunction - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC2685818/](https://pmc.ncbi.nlm.nih.gov/articles/PMC2685818/) 173. Neural coding of visual objects rapidly reconfigures to reflect sub-trial shifts in attentional focus | bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2021.05.25.445712v1.full-text](https://www.biorxiv.org/content/10.1101/2021.05.25.445712v1.full-text) 174. (PDF) Consciousness in the universe: a review of the 'ORCH OR' theory - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/257134660_Consciousness_in_the_universe_a_review_of_the_'ORCH_OR'_theory](https://www.researchgate.net/publication/257134660_Consciousness_in_the_universe_a_review_of_the_'ORCH_OR'_theory) 175. Orchestrated objective reduction - Wikipedia, accessed April 20, 2025, [https://en.wikipedia.org/wiki/Orchestrated_objective_reduction](https://en.wikipedia.org/wiki/Orchestrated_objective_reduction) 176. 14 CONSCIOUSNESS IN THE UNIVERSE AN UPDATED REVIEW OF THE “ORCH OR” THEORY - Anesthesiology, accessed April 20, 2025, [https://anesth.medicine.arizona.edu/sites/default/files/hameroff-penrose_updated_review_of_orch_or_2016_b2237_ch-14_revised-2-3.pdf](https://anesth.medicine.arizona.edu/sites/default/files/hameroff-penrose_updated_review_of_orch_or_2016_b2237_ch-14_revised-2-3.pdf) 177. Quantum computation in brain microtubules? The Penrose-Hameroff "Orch OR" model of consciousness - Esalq, accessed April 20, 2025, [http://www.esalq.usp.br/lepse/imgs/conteudo_thumb/Quantum-computation-in-brain-microtubules-The-Penrose-Hameroff-Orch-or-model-of-consciousness.pdf](http://www.esalq.usp.br/lepse/imgs/conteudo_thumb/Quantum-computation-in-brain-microtubules-The-Penrose-Hameroff-Orch-or-model-of-consciousness.pdf) 178. 'Orch OR' is the most complete, and most easily falsifiable theory of consciousness, accessed April 20, 2025, [https://www.tandfonline.com/doi/full/10.1080/17588928.2020.1839037](https://www.tandfonline.com/doi/full/10.1080/17588928.2020.1839037) 179. Consciousness, Cognition and the Neuronal Cytoskeleton – A New Paradigm Needed in Neuroscience - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC9245524/](https://pmc.ncbi.nlm.nih.gov/articles/PMC9245524/) 180. Discovery of quantum vibrations in 'microtubules' inside brain ..., accessed April 20, 2025, [https://www.sciencedaily.com/releases/2014/01/140116085105.htm](https://www.sciencedaily.com/releases/2014/01/140116085105.htm) 181. Consciousness in the universe: a review of the 'Orch OR' theory - PubMed, accessed April 20, 2025, [https://pubmed.ncbi.nlm.nih.gov/24070914/](https://pubmed.ncbi.nlm.nih.gov/24070914/) 182. A Further Review of 'Orch OR' Theory: The Universe in Consciousness - ResearchGate, accessed April 20, 2025, [https://www.researchgate.net/publication/281229377_A_Further_Review_of_'Orch_OR'_Theory_The_Universe_in_Consciousness](https://www.researchgate.net/publication/281229377_A_Further_Review_of_'Orch_OR'_Theory_The_Universe_in_Consciousness) 183. Empirical Validation of the Orchestrated Objective Reduction (Orch-OR) Theory: Integrating - OSF, accessed April 20, 2025, [https://osf.io/5v7dp/download](https://osf.io/5v7dp/download) 184. Quantum Models of Consciousness from a Quantum Information Science Perspective - arXiv, accessed April 20, 2025, [https://arxiv.org/html/2501.03241v1](https://arxiv.org/html/2501.03241v1) 185. Beyond Trial-Based Paradigms: Continuous Behavior, Ongoing ..., accessed April 20, 2025, [https://www.jneurosci.org/content/38/35/7551](https://www.jneurosci.org/content/38/35/7551) 186. Quantifying decision-making in dynamic, continuously evolving environments - eLife, accessed April 20, 2025, [https://elifesciences.org/articles/82823](https://elifesciences.org/articles/82823) 187. Neural timescales from a computational perspective arXiv:2409.02684v1 [q-bio.NC] 4 Sep 2024, accessed April 20, 2025, [https://arxiv.org/pdf/2409.02684](https://arxiv.org/pdf/2409.02684) 188. The Discrete and Continuous Brain: From Decisions to Movement ..., accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6115199/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6115199/) 189. Goal-Directed Decision Making with Spiking Neurons | Journal of Neuroscience, accessed April 20, 2025, [https://www.jneurosci.org/content/36/5/1529](https://www.jneurosci.org/content/36/5/1529) 190. Decisions in Changing Conditions: The Urgency-Gating Model ..., accessed April 20, 2025, [https://www.jneurosci.org/content/29/37/11560](https://www.jneurosci.org/content/29/37/11560) 191. A general recurrent state space framework for modeling neural dynamics during decision-making - Proceedings of Machine Learning Research, accessed April 20, 2025, [http://proceedings.mlr.press/v119/zoltowski20a/zoltowski20a.pdf](http://proceedings.mlr.press/v119/zoltowski20a/zoltowski20a.pdf) 192. Neural underpinnings of the evidence accumulator - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC5777584/](https://pmc.ncbi.nlm.nih.gov/articles/PMC5777584/) 193. A Bayesian Attractor Model for Perceptual Decision Making | PLOS ..., accessed April 20, 2025, [https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1004442](https://journals.plos.org/ploscompbiol/article?id=10.1371/journal.pcbi.1004442) 194. Modelling decision-making biases - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2023.1222924/pdf](https://www.frontiersin.org/journals/computational-neuroscience/articles/10.3389/fncom.2023.1222924/pdf) 195. Neural dynamics and circuit mechanisms of decision-making - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC4065788/](https://pmc.ncbi.nlm.nih.gov/articles/PMC4065788/) 196. Neural circuit models for evidence accumulation through choice ..., accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2023.09.01.555612v4.full-text](https://www.biorxiv.org/content/10.1101/2023.09.01.555612v4.full-text) 197. Attractor dynamics reflect decision confidence in macaque prefrontal cortex - bioRxiv, accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2023.09.17.558139v1.full-text](https://www.biorxiv.org/content/10.1101/2023.09.17.558139v1.full-text) 198. Neural dynamics of reversal learning in the prefrontal cortex and recurrent neural networks, accessed April 20, 2025, [https://elifesciences.org/reviewed-preprints/103660v1](https://elifesciences.org/reviewed-preprints/103660v1) 199. Sustained neural rhythms reveal endogenous oscillations supporting speech perception - PMC - PubMed Central, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC7946281/](https://pmc.ncbi.nlm.nih.gov/articles/PMC7946281/) 200. Entrainment of Perceptually Relevant Brain Oscillations by Non-Invasive Rhythmic Stimulation of the Human Brain - Frontiers, accessed April 20, 2025, [https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2011.00170/full](https://www.frontiersin.org/journals/psychology/articles/10.3389/fpsyg.2011.00170/full) 201. Phase-tuned neuronal firing encodes human contextual representations for navigational goals - PMC, accessed April 20, 2025, [https://pmc.ncbi.nlm.nih.gov/articles/PMC6040884/](https://pmc.ncbi.nlm.nih.gov/articles/PMC6040884/) 202. An Oscillatory Hierarchy Controlling Neuronal Excitability and ..., accessed April 20, 2025, [https://journals.physiology.org/doi/full/10.1152/jn.00263.2005](https://journals.physiology.org/doi/full/10.1152/jn.00263.2005) 203. Does the phase of ongoing EEG oscillations predict auditory ..., accessed April 20, 2025, [https://www.biorxiv.org/content/10.1101/2020.07.17.209387v3.full](https://www.biorxiv.org/content/10.1101/2020.07.17.209387v3.full) 204. EEG, MEG and neuromodulatory approaches to explore cognition - Zurich Open Repository and Archive, accessed April 20, 2025, [https://www.zora.uzh.ch/id/eprint/221125/6/1_s2.0_S0278262620302803_main.pdf](https://www.zora.uzh.ch/id/eprint/221125/6/1_s2.0_S0278262620302803_main.pdf) 205. Individual Neurons in the Cingulate Cortex Encode Action Monitoring, Not Selection, during Adaptive Decision-Making | Journal of Neuroscience, accessed April 20, 2025, [https://www.jneurosci.org/content/39/34/6668](https://www.jneurosci.org/content/39/34/6668) 206. Signatures of time interval reproduction in the human electroencephalogram (EEG) | Imaging Neuroscience - MIT Press Direct, accessed April 20, 2025, [https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00279/124045/Signatures-of-time-interval-reproduction-in-the](https://direct.mit.edu/imag/article/doi/10.1162/imag_a_00279/124045/Signatures-of-time-interval-reproduction-in-the)