Papers

Foundation models are beginning to reshape brain-signal analysis by moving the field beyond task-specific decoding pipelines toward reusable models pretrained on broad neural datasets. Magnetoencephalography (MEG) is a compelling but still underdeveloped target for this shift: it captures human cortical dynamics at millisecond resolution while offering stronger spatial interpretability than EEG, making it especially valuable for source-resolved studies of perception, language, cognition, and clinical brain function. Yet MEG foundation models remain at an early stage, with only a small number of MEG-specific and MEG-inclusive multi-modal models, modest pretraining corpora, and emerging but still limited benchmarks. This perspective lays down the basic concepts needed to understand MEG foundation models and provides a didactic overview of the field's key design choices, including tokenization, sensor- versus source-space representations, sensor-geometry encoding, backbone architectures, self-supervised objectives, and pretraining data. We then offer a roadmap for future development, organized around native MEG pretraining, adaptation of EEG foundation models, transfer from generic time-series models, and multi-modal integration with EEG, fMRI, MRI, behaviour, and stimulus features. We highlight the need for coordinated infrastructure, including diverse and reusable MEG datasets, rigorous evaluation across subjects, sites, tasks, and clinical settings, and responsible data-sharing practices that address consent, privacy, access, and governance.

Cortical neurons fire sparsely – often fewer than one spike per sensory window – making rate coding insufficient and temporal coding a necessity. That conduction delays convert firing order into synchrony is long established. What governs which class of temporal feature a neuron detects – one volley of coincident input, or two in a particular order – has not been examined. We propose a delay-signature framework in which the axonal conduction delays converging on a dendritic branch constitute a physical key: only input sequences whose spike-time differences the delays compensate arrive synchronously, and coincidence detection, via calcium plateau thresholds, converts that synchrony into an all-or-none output. In simulations of an integrator-neuron model we report three results. First, a single physical scalar – the dispersion of the delay set – moves a population from event detection to order-selective sequence detection. The transition is emergent under random delays and connectivity: at narrow dispersion sequence detectors do not exist, and the dispersion at which they overtake event detectors tracks the inter-event interval with a slope statistically indistinguishable from one. This maps a computational distinction onto the anatomical one between myelinated and unmyelinated projections, making myelination a switch on what a neuron computes, not only a regulator of speed. Second, the same dispersion sets the code's limits: it bounds the longest codable interval and fixes an absolute timing tolerance of about a millisecond, with slowing better tolerated than speeding. Third, that millisecond window and horizontal conduction velocity together predict cortical column diameter, and the two areas with direct measurements fall where the relation puts them. One anatomically measurable parameter thus sets what a neuron detects and the limits of what it can represent.

Temporal integration gives continuous-time recurrent networks memory, but in deep stacks it also delays bottom-up signals and attenuates top-down errors. We develop Recursive Quadrature Filters (RQFs), biologically motivated complex-valued temporal filters that are a special case of diagonal state-space models (SSMs), and ask whether this failure mode can be addressed by making each layer's bottom-up input prospective. Starting from an energy model, we derive the RQF dynamics and show that each RQF is a band-pass filter whose learnable parameters control its tuning frequency and bandwidth. We then make each layer's bottom-up input prospective using a parameter-free two-tap update that leaves the recurrent transition and parallel scan unchanged. We extend this correction to general diagonal SSMs and show that it mitigates depth-dependent gradient attenuation when temporal gradients are truncated, i.e., spatial-only backpropagation. We evaluate the intervention in RQFs, S5, and ORGaNICs (a nonlinear gated RNN) trained using full backpropagation through time (BPTT) and spatial-only backpropagation. Under full BPTT, prospective variants match or outperform their non-prospective controls in every model and configuration. A non-residual width-32 six-layer RQF reaches 96.09% accuracy on raw-audio Speech Commands with 31.9k parameters; a width-64 six-layer RQF reaches 83.56% on the 16,384-step Path-X task. These results identify RQFs as a parameter-efficient recurrent substrate and prospective-input coding as an input-side correction for deep continuous-time recurrent networks.

Here, we report high-order functional network connectivity as a promising way for studying the brain connectome. Traditional functional connectivity approaches capture only pairwise relationships between brain regions, overlooking complex multivariate dependencies that underlie cognition and behavior. First, we demonstrated that high-order interactions capture more information and can distinguish between resting-state and task-state brain activity. Second, we introduce a matrix-based entropy-functional method for estimating triadic interactions, which are statistical dependencies among triplets of brain regions, and apply it to large-scale functional brain networks. The resulting triadic networks revealed distinct community patterns that complement those observed in traditional pairwise functional connectivity analyses and simultaneously capture additional connection information. Despite the potential combinatorial explosion of triadic configurations, the networks exhibited constrained and hierarchical structures that allowed computation and interpretation. These findings position triadic connectivity as a promising next-step functional connectivity framework for probing brain network organization and high-order neural interactions, while also highlighting key biological and technical challenges that require careful consideration.

Creatures that display 'hedonic place preference behaviour' are thought by many scientists to experience feelings, on the assumption that their attraction to pleasure-producing substances which lack nutritional value (e.g. cocaine, morphine) cannot easily be attributed to unconscious instinctual behaviour. In this paper, we discuss how a simple artificial agent that instantiates attributes of an affective system engaging in felt uncertainty about its intrinsic needs in relation to environmental resources can similarly display hedonic place preference behaviour – through an apparently subjective form of information processing – while simultaneously being entirely deter-ministic. We outline some implications of this artificially engineered behaviour for our understanding of the physical basis of consciousness and the experience of free will.

We present a unified computational approach to tensor-based morphometry in detecting the brain surface shape differences between two clinical groups based on magnetic resonance images. Our approach is novel in a sense that we combined surface modeling, surface data smoothing and statistical analysis in a coherent unified mathematical framework. The cerebral cortex has the topology of a 2D highly convoluted sheet. Between two different clinical groups, the local surface area and curvature of the cortex may differ. It is highly likely that such surface shape differences are not uniform over the whole cortex. By computing how such surface metrics differ, the regions of the most rapid structural differences can be localized. To increase the signal to noise ratio, diffusion smoothing based on the explicit estimation of Laplace-Beltrami operator has been developed and applied to the surface metrics. As an illustration, we demonstrate how this new tensor-based surface morphometry can be applied in localizing the cortical regions of the gray matter tissue growth and loss in the brain images longitudinally collected in the group of children.

Speech brain-computer interfaces (speech BCIs) translate neural activity into language, offering a path towards restoring speech for people with paralysis and, more broadly, enabling new forms of natural human-computer interaction. Despite this promise, the field lacks a common measure of progress because systems use different datasets, recording methods, types of speech, and vocabularies, so their reported scores are rarely comparable. Underlying this measurement problem are two unresolved questions: (i) what distribution of words should a speech BCI enable a user to communicate, and (ii) how much information from this distribution can a system convey. We address both by deriving open-vocabulary mutual information (OVMI), an information-theoretic quantity that measures the information conveyed by a decoder relative to a reference distribution over the words a user may wish to communicate. This allows capabilities measured under different conditions, such as distinct vocabularies, to be evaluated on a common communication scale. We show that ordinarily reported accuracy, word error rate (WER), and other metrics computed only over the words a system supports can overstate how much of a user's intended speech the system can communicate. We then use OVMI to compare existing systems, expose trade-offs between how much of the user's language a system supports and how accurately it decodes those words, show that these comparisons depend on what the user is expected to communicate, and demonstrate that selecting a vocabulary to maximise OVMI yields up to 16.3% relative improvement in accuracy across three speech domains. OVMI therefore provides the speech BCI community with a principled way to compare heterogeneous systems, improve vocabulary design, and measure progress in the field.

The brain is highly proficient at making sense of noisy and ambiguous sensory inputs. Predictive processing hypothesises that this ability relies on prediction. However, it is unclear why the brain would have evolved to predict the sensory world, a computationally expensive process, in order to aid perception. Here we use simulations to argue that prediction naturally emerges in systems optimised for perception. We train recurrent neural networks (RNNs) to denoise a tokenised version of Bach's compositions at a range of noise levels. Afterwards, we enquire whether the states of the networks contain predictive information about the next token. We test this by freezing the RNN weights and training a linear readout from its states on prediction. We compare the performance of the linear readout with that of an independently trained linear benchmark model. The results show that the linear readout from the RNNs outperforms the benchmark model at moderate levels of noise, indicating that the networks rely on a predictive mechanism to support perception. We further show that the responses of the RNNs to sensory inputs are proportional to prediction error. Together, the results demonstrate that neural signatures of predictive processing emerge, without any further training constraints, from optimisation of perception.

This paper argues that fungal mycelial networks exhibit minimal cognition through memory-integrated adaptive regulation. Drawing on cybernetic and enactivist frameworks, I develop a non-representational account of memory as the organism's capacity to modulate behavior based on temporally extended environmental coupling. I propose four operational criteria for minimal cognition: feedback-guided regulation of behavior, maintenance of internal viability conditions, structural modulation based on past environmental interactions, and plasticity across time scales that supports anticipatory adaptivity. Empirical evidence demonstrates that fungi meet all four criteria through distributed memory mechanisms: fungal networks exhibit directional regrowth toward previously encountered resources even after spatial displacement; stress priming persists across multiple cell divisions; Spitzenkörper-mediated directional persistence in constrained environments; and transgenerational memory through spore imprinting. These findings challenge representationalist assumptions in cognitive science by showing that memory and cognition can emerge from morphodynamic, biochemical, and electrophysiological processes without necessary neural substrates or symbolic representations. Fungal cognition demonstrates that the organizational principles underlying cognition - feedback-driven adaptation, norm-preservation, and historical coupling - can be realized in radically different material substrates, expanding our understanding of what counts as a cognitive system.

Digital twin technologies could transform neuroscience and biomedicine by creating predictive computational representations of living organisms. However, most animal digital twins model neural circuits, anatomy, or biomechanics separately rather than integrating the processes that generate behavior. We argue that animal digital twins should instead be conceived as embodied dynamical systems that unify neural activity, body mechanics, sensory feedback, and environmental interactions. We propose a neuro-musculoskeletal digital twin of the mouse that combines multimodal anatomical reconstruction from X-ray CT, high-resolution white-light sections, and Scx-GFP imaging with biomechanical simulation, Bonhoeffer–van der Pol neural dynamics, and tactile feedback. This framework forms a closed sensorimotor loop in which behavior emerges through continuous interactions among the nervous system, musculoskeletal system, and environment. The Bonhoeffer–van der Pol model provides a computationally tractable dynamical foundation for large-scale simulation of these interactions. Neuro-musculoskeletal digital twins could provide a convergence point for computational neuroscience, biomechanics, artificial intelligence, and robotics. When coupled with adaptive learning and autonomous experimentation, they may develop from passive simulations into active scientific instruments that generate hypotheses, predict interventions, and guide experiments. Such embodied digital twins could advance the study of biological intelligence and support new adaptive biomedical and robotic systems.

This chapter reconstructs the Hopfield network as a physical theory of memory rather than merely an early neural-network algorithm. It begins with the problem as it stood before 1982-threshold logic, Hebbian association, correlation memories, and recurrent binary networks-and isolates what Hopfield's synthesis added: a dynamical definition of content-addressable memory, a symmetric recurrent architecture with a Lyapunov function, a Hebbian embedding of patterns in its couplings, and a physical account of basins, robustness, and graceful degradation. The binary and graded-response energy functions are derived in full, together with the signal-crosstalk decomposition governing pattern stability, the mean-field theory of retrieval at extensive load, and the zero-temperature retrieval spinodal at (alpha 0.138) established by Amit, Gutfreund, and Sompolinsky. The energy-based program is then followed through analog optimization networks, polynomial dense associative memories, exponential interactions, and modern continuous Hopfield updates, including the precise conditions under which the update becomes scaled dot-product attention. Throughout, capacity claims are tied to their disorder ensemble, scaling limit, and success criterion, showing why numerically different storage limits need not conflict. A closing assessment distinguishes established results from surviving principles, assumption-bound limitations, and open problems, treating the Hopfield network as an effective theory whose symmetry, locality, and point-neuron assumptions delimit its biological reach. Fixed-seed numerical experiments expose the mechanisms discussed but do not substitute for analytical results.

This chapter reconstructs the McCulloch-Pitts program as a physics of neural computation rather than the familiar cartoon of a binary neuron. The 1943 logical calculus is developed in both directions: given a net, characterize the propositions realized by its activity; given an admissible logical expression, construct a net that realizes it. We recover the original distinction between thresholded excitatory summation and absolute inhibitory veto-one the weighted-threshold form cannot preserve for arbitrarily large excitatory inputs-and read unit-time delay as the physical realization of logical depth. Recurrence is treated exactly: an autonomous, deterministic network of finitely many binary units has a finite state space, so every trajectory eventually enters a periodic orbit-a fact about finite-state dynamics, not unbounded Turing computation. A single threshold element realizes only linearly separable Boolean functions, whereas finite feedforward networks of them synthesize any Boolean function on a finite domain. We then follows McCulloch and Pitts beyond threshold logic. The 1945 heterarchy paper turns cyclic preference into an obstruction to representation by a scalar utility. The 1947 work on universals asks how a physical network can identify inputs related by nuisance transformations, developed here via group averaging and feedback canonicalization. The 1959 frog-retina study makes the adequate-stimulus question experimental, revealing parallel invariant operations before the brain proper. Spike-triggered analysis shows how a nonlinearly driven neuron can have a vanishing first-order average while second-order statistics recover its hidden selectivity: methodological failure can masquerade as physiological absence. Modern mathematical tools are used without projecting their notation onto the historical papers, and limitations of the idealization are stated explicitly.

The pulling illusion induced by asymmetric vibration stimuli has attracted attention for its potential applications in rehabilitation and sensory assessment. However, the underlying mechanism of the pulling illusion remains unclear. This study addressed the central question of whether peripheral vibrotactile sensitivity alone is sufficient for the illusion to emerge or whether processing beyond basic vibration detection is also required. Neurological disorders can involve impairments at different levels of the nervous system, providing an opportunity to examine this question. Accordingly, we evaluated directional discrimination performance for the pulling illusion and fingertip vibration detection thresholds in 25 participants with diverse neurological disorders affecting different levels of the nervous system, from peripheral to central. Clustering analysis identified contrasting profiles, with high directional discrimination performance despite elevated vibration detection thresholds and chance-level performance despite relatively low-to-intermediate thresholds. In the generalized linear mixed model, motor-related signs, including hemiplegia and tremor, showed a robust negative association with directional discrimination performance, whereas vibration detection threshold was not robustly associated with performance. Furthermore, in participants with hemiplegia, directional discrimination performance was around chance level on the affected side and close to 100% on the unaffected side, despite stimulus amplitudes well above the measured vibration detection thresholds on both sides. Collectively, these findings suggest that the pulling illusion depends on perceptual processing beyond basic vibration detection, through which asymmetric vibration is experienced as directional pulling.

There is growing interest in the proposed use of EEG signals as biometric credentials, but thus far there has been little research on the reliability and security of such biometrics. Prior adversarial tests have focused on deep-learning classifiers and assumed attackers have full access to the classifier model. This has left unexamined other, more popular categories of neural signature methods as well as the more realistic case of an adversary having only black-box access to a classifier. In this paper we develop a collection of adaptive attack algorithms which learn to fool an authentication system via targeted alterations of stolen EEG recordings, without requiring any knowledge of the authentication system itself. Tested on 6 public datasets spanning three recording conditions (reacting to visual stimuli, imagining hand movements, and resting), it reveals that different signaturing approaches vary significantly in their degrees of vulnerability to adversarial attacks. We show that vulnerability to spoofing attack is greatly impacted by the recording conditions, with significant variation depending on task at time of recording. Finally, we provide recommendations for improving neural signature biometrics based on the results of our adversarial testing.

Patients with depression present with diverse symptom profiles, yet clinical practice routinely reduces this variation to a single severity score. Large language models (LLMs) can potentially capture various symptoms and their severity from patient speech. However, how depressive symptoms are represented inside LLMs remains poorly understood, limiting clinical trust. To examine whether internal model activations match clinician judgment, we analyzed the residual stream of Gemma-3-27B-PT using mechanistic interpretability techniques. Recording activations across symptom descriptions drawn from validated clinical instruments, we found that symptom groups geometrically separated the most at layer 21 across multiple distance metrics. Using Semantic Projection, we then projected held-out naturalistic text onto Symptom Vectors constructed from these instruments. The resulting per-symptom coefficients preserved clinician-annotated rank ordering across mood, somatic, and suicidality axes. Furthermore, a single depression vector in Layer 21 separates held-out depressive from non-depressive text (AUC = 0.789), which can be used as an emotional valence gate that restricts symptom projection to depressive speech. These results reveal a decorrelated, clinician-aligned symptom signal readable directly from internal activations, offering a mechanistic foundation for interpretable depression-assessment tools.

Artificial intelligence (AI) is transforming brain-computer interfaces (BCIs) from task-specific neural decoders into adaptive systems that complete language, smooth movement, regulate rehabilitation support and adjust stimulation. These capabilities can increase speed, fluency, usability and clinical reach, yet conventional performance metrics may overlook losses in intent fidelity, authorship, agency, therapeutic challenge and durable clinical benefit. I define neuroadaptive overfitting as a closed-loop failure mode in which an AI-mediated BCI becomes over-optimized to short-term proxies of success, including reduced effort, rapid acceptance, lower workload or smooth task completion, while drifting from the user's durable goals. I then propose Slow-Fast BCI, a framework for pacing AI assistance according to decoder evidence, uncertainty, contextual and clinical stakes, fatigue, and user- or clinician-defined goals. The framework distinguishes fast assistance when intent is clear and stakes are low, guarded assistance under uncertainty and slow assistance when misalignment could compromise safety, agency, authorship, motor learning or therapeutic value. Across communication, motor-control, neurorehabilitation and closed-loop neuromodulation applications, I outline corresponding safeguards and evaluation measures. This Perspective argues that AI-mediated BCIs should be evaluated not only by decoding accuracy and task performance, but also by how AI assistance is deployed: when systems act autonomously, seek confirmation, preserve user effort or return control to the user.

Here we present the Active Visual Semantics (AVS) dataset, a large-scale collection of magnetoencephalography (MEG) and eye-tracking data recorded while five participants freely explored 4,080 natural scenes over 10 sessions each, yielding more than 200,000 fixation epochs in total. Unlike existing neuroimaging datasets that rely on passive viewing with enforced central fixation, AVS captures brain activity during active scene exploration, including self-generated saccades and fixations. A semantic captioning task on 25% of the trials provides behavioural measures linking gaze to scene understanding and memory. In addition to neural and behavioural data, AVS includes per-fixation object category labels, human-rated annotations of the appearance of fixation targets in the scene captioning task and pupil dynamics. Individual head stabilisation casts were used during MEG data collection, which alongside with structural MRI scans, enabled precise cross-session source reconstruction. Using artificial neural network (ANN) encoding models we demonstrate that individual fixation-aligned MEG epochs hold visual content-specific signal, despite the challenges that active scene viewing poses for MEG signal quality. Further, we use fixation-aligned representation similarity analysis (RSA) and demonstrate that we can derive fixation object category averages that yield representational geometries which are highly reliable across participants. Both in MEG sensor and source space this structure is validated by its robust alignment with ANN object-level representational geometry. Taken together, AVS provides a rich resource for investigating a large variety of questions regarding the neural mechanisms of active vision, object recognition and scene captioning during natural viewing, and the relationship between gaze behaviour and memory encoding.

Geometric eigenmodes provide a compact and biologically grounded representation of large-scale neural activity. Previous work demonstrated that they can mitigate the underdetermined nature of electroencephalographic (EEG) and magnetoencephalographic (MEG) source localisation, an ill-posed inverse problem in which neural activity is reconstructed from non-invasive recordings. Beyond their spatial structure, neural field theory predicts the temporal evolution of eigenmodes through analytically derived transfer functions. Motivated by this framework, the present work investigates whether these transfer functions can be used to introduce temporal constraints into EEG source imaging. The approach is evaluated using simulated seizure dynamics generated by coupled Epileptor neural mass models. Transfer functions derived directly from neural field theory were found to be generally ineffective as temporal constraints for source localisation, primarily because they neglect cross-eigenmode coupling. Incorporating empirically estimated coupling terms substantially improves localisation performance, particularly in noisy conditions. Although estimating these eigenmode coupling interactions from experimental data remains challenging, the findings motivate dynamical source imaging approaches that combine spatial eigenmode structure with empirically informed cross-modal dynamics.

The corticothalamic circuit supports rhythms with timescales that differ by orders of magnitude: sleep spindles, the sigma-band events of non-rapid-eye-movement (NREM) sleep, and infra-slow fluctuations near 0.02Hz that organize when spindles occur. Because the anatomy is the same in both cases, architecture alone cannot determine which rhythm the circuit expresses. We ask whether the temporal structure of the circuit's own feedback can. In a four-population Wilson–Cowan model comprising cortical excitatory and inhibitory populations, thalamic relay cells, and the thalamic reticular nucleus (TRN), we first establish how connectivity controls access to oscillatory behavior, and then introduce temporal coupling as either a weak Gamma distributed delay or a discrete delay. We investigate three distinct connectivity levels: recurrent cortical excitation gates whether the circuit can oscillate at all, the reciprocal relay-TRN pair determines where the oscillation lies and how it is configured, sustained, and terminated, and reticular self-inhibition limits its extent. We then examine how these connectivity-dependent regimes are affected by delayed coupling. Although delay does not change the equilibria themselves, it can substantially alter their stability and the organization of the resulting oscillatory dynamics. Under weak Gamma integration, short delays support spindle-compatible oscillations in the sigma band, while longer delays give rise to a much slower regime near 0.02Hz. The discrete-delay formulation produces a qualitatively different and more complex bifurcation structure. Together, these results show that the dynamics of the corticothalamic circuit depend not only on its connectivity, but also on the temporal organization of interactions within the circuit.

Canonical neural circuit motifs are usually described functionally: divisive normalization rescales population activity by a pooled signal, and winner-take-all competition selects one pattern through recurrent excitation and shared inhibition. We represent them, and their compositions, algebraically as finite transformation systems and analyze the transition monoids generated by their input-conditioned updates, distinguishing structure already present in a generator from structure that appears only through composition, and, on a joint state space, structure inherited from one factor from structure that lives on a joint configuration. Individually aperiodic updates can generate non-aperiodic monoids. In the WTA, every frozen-drive generator collapses to fixed points, yet short input sequences create local cycles of winner-dependent inhibitory gating: globally dissipative dynamics with a reversible action. The strongest result arises in WTA-to-DN composition. The composed monoid then contains a genuinely composite local cycle in which normalization state and the winner's gating state change together, although every primitive generator is aperiodic. Holonomy analysis certifies this as a group component of the Krohn-Rhodes cascade rather than an incidental cycle, and finds most group-carrying image sets on joint configurations, whereas the uncoupled product has none. An exhaustive interface sweep shows that the composite cycle is a property of the coupling rather than of a chosen map. If motifs are building blocks of neural computation, composing them is a form of programming: one chooses primitives and interfaces so that the generated algebra has the intended repertoire. The transition monoid is that repertoire - what a primitive presents to any later construction. Recurrent circuits are compositional transformation systems; their algebra constrains what they can be programmed to compute.

Neural correlates of spatial cognitive map are well documented, yet exactly how neural circuits perform spatial navigation in complex environments - e.g., reaching a goal while avoiding obstacles - remains largely unclear. Here, we show that a hippocampal network with appropriate recurrent connections can naturally achieve optimal goal-directed navigation via its relaxation dynamics. Specifically, we consider that the recurrent weights between the neurons represent the transition probabilities between spatial locations encoded by neurons; obstacles such as walls and blocked corridors are therefore reflected by the vanishing of connection weights. This connection pattern can be learned in the hippocampus via behavioral-timescale synaptic plasticity (BTSP) while the animal is exploring the environment. When a goal signal is presented, the network dynamics will relax into an activity field representing the goal location. We prove that this field is mathematically equivalent to the desirability field of a Linearly-solvable Markov Decision Process (LMDP), and the local log-gradient of the field indicates the navigation direction. Both theoretical analyses and simulations demonstrate that this recurrent network dynamics-mediated navigation is efficient and robust in environments with complex obstacle layouts. Moreover, only low-rank updates of the network's connection pattern are needed when the environment has local changes. We hope this study offers insight into a general circuit principle for planning in abstract rational maps in the brain beyond spatial navigation.

Disordered dynamical systems comprising many interacting units, from ecological communities to neural circuits, are ubiquitous, and understanding how connectivity shapes their collective behavior is a central theoretical challenge. One long-recognized feature of neural circuits is local connectivity balance, in which the excitatory and inhibitory weights converging onto each unit approximately cancel. Although local connectivity balance has been proposed to serve functions such as gating incoming signals, its effect on collective network dynamics remains unclear. Here we analytically study randomly connected recurrent networks with varying degrees of local connectivity balance. We show that this balance leaves the connectivity spectrum unchanged yet drastically reshapes the dynamics in a manner that depends critically on the single-unit nonlinearity. Local balance suppresses unbounded growth of the network state and stabilizes network dynamics when the activation function scales linearly or faster, whereas it drives the network into chaos when the activation function is sub-linear or saturating. Importantly, these effects vanish for odd activation functions, which are commonly assumed in previous work. We further find that, for saturating nonlinearities, the effective dimension of the dynamics varies nonmonotonically with the degree of balance. We show that all these phenomena arise from a unifying mechanism: the suppression of a self-generated feedback input by local connectivity balance. Our results identify local connectivity balance as a previously overlooked control parameter for collective dynamics in realistic disordered networks.

We propose a neurobiologically plausible model of cognition that combines the advantages of connectionist and symbolic systems, and that can explain a wide range of cognitive phenomena. This model, called Rate-Coding Bundle Memory (RCBM), is based on the Symbolic Subsystem Hypothesis, which posits that the brain implements a symbolic subsystem within its fundamentally connectionist nature. RCBM is a hybrid model that uses rate coding to represent symbols in a continuous space, and it uses a bundle memory system to store and retrieve these symbols. The model is capable of solving a wide range of cognitive phenomena, including one-shot learning, pattern separation, and the binding problem. We argue that RCBM provides a promising framework for understanding the nature of cognition, and that it can be used to develop more sophisticated models of cognition in the future.

40-Hz ASSR is reduced in schizophrenia, but it is unclear if this reflects altered auditory input or cortical E/I dynamics. We hypothesized that similar group differences could arise via distinct model mechanisms. EEG gamma% and ITPC from 21 HC and 21 SCZ constrained an auditory front-end coupled to a Wilson-Cowan E/I model. We compared front-end-restricted, back-end-restricted, and full-joint parameter searches, plus perturbation and fixed-point analyses. HC means exceeded SCZ for both metrics (not individually significant). All three models reproduced HC>SCZ but located group differences differently: front-end via input transformation, back-end via cortical dynamics, full-joint via both. The full-joint solution was most robust to perturbation. Fixed-point analysis revealed similar outputs with distinct local dynamics. All models reproduced the HC>SCZ pattern, suggesting schizophrenia pathophysiology may involve altered sensory encoding, altered cortical E/I, or both. This framework enables future patient-level mechanistic comparison and, after validation, may support individualized stratification.

Cocaine Use Disorder (CUD) is associated with widespread alterations in large-scale functional brain networks, yet the mechanisms contributing to these changes and their relationship to clinical and cognitive outcomes remain poorly understood. To address this gap, we introduce a framework to extract structurally informed dynamic functional connectivity patterns. We then leverage these connectivity patterns to characterize differences in functional brain network organization associated with CUD and to examine their relationship with clinical measures. Specifically, we applied Laplacian spectral smoothing to each participant's functional connectivity matrix using individualized structural priors derived from diffusion imaging. These structurally informed connectivity features were subsequently used to examine cross-network interactions and characterize dynamic community organization across functional brain states. Our findings indicate that individuals with cocaine use disorder exhibit increased integration and recruitment accompanied by reduced flexibility in the functional brain networks, with the most pronounced alterations in visual, attentional, and control systems. In addition, structurally informed functional connectivity features were predictive of weekly cocaine use within the CUD cohort. Overall, these results highlight the value of structurally informed dynamic connectivity measures for characterizing network-level alterations associated with cocaine addiction and for linking these alterations to clinically meaningful measures of cocaine use severity.

Vision neuroscience has experienced a surge in the collection and use of large-scale datasets of brain responses to naturalistic images. However, static images lack the temporal dimension essential for understanding how vision is solved in the brain during dynamic real life settings. To facilitate the study of the neural correlates of dynamic visual event perception, we introduce the EEG Moments Dataset (EMD). EMD consists of 128-channel EEG responses and eye-tracking recordings of 6 human participants viewing 1,102 short naturalistic videos (3-second long; with audio track) while maintaining central fixation. We show that EMD's EEG responses well encode stimulus-related information, exhibit a temporal correspondence with the video stimuli, and have a rich representational content revealed by brain encoding models based on different feature spaces. Furthermore, complemented by the BOLD Moments Dataset (BMD) - an existing large-scale dataset of human functional magnetic resonance imaging (fMRI) responses for the same videos - EMD enables spatio-temporally resolved investigations of brain responses to dynamic visual events. We release EMD's EEG and eye-tracking data in both raw and preprocessed format, along with the 1,102 video stimuli, and rich stimulus metadata. Finally, we provide an interactive code tutorial to familiarize with EMD's preprocessed data, stimuli, and stimulus metadata.

Deep neural networks are high-dimensional dynamical systems whose function depends on the stable propagation of activity and perturbations across many layers. Maintaining suitable dynamical regimes may therefore be essential for robust learning and for preventing dynamical instabilities during training. Here, we show that the global dynamical state of a deep neural network can be autonomously regulated by purely local homeostatic plasticity. Neuronal activity is inferred from responses across inputs, and individual synapses are strengthened or weakened using only the activity of their postsynaptic neuron. Without measuring any global network property, this rule drives networks from both subcritical and supercritical initial conditions toward a common critical state, characterized by conserved activity propagation and a vanishing largest finite-time Lyapunov exponent. When combined with gradient-based learning, homeostatic adaptation counteracts the training-induced drift toward supercritical dynamics, while revealing a competition between dynamical regulation and task optimization. Our results demonstrate how adaptive self-organization can be implemented in deep neural networks and how local plasticity can control their collective dynamical operating point.

Linking the internal representations of deep neural networks (DNNs) to human mental representations is important for using DNNs as computational models of human vision. Existing DNN representations remain insufficiently similar to human mental representations, which are not directly observable and are therefore commonly measured through large-scale similarity judgments of object images. A natural approach to narrowing this gap is to directly transfer the relational structure of human representations into DNNs, and previous studies have reported improved human-DNN representational similarity. However, whether this improvement holds under stricter evaluation remains untested in two respects: fine-grained alignment at the individual-object level, and generalization to a human embedding derived from a dataset independent of the training data. Here, we employ an unsupervised comparison method, Gromov-Wasserstein optimal transport (GWOT), which estimates human-DNN correspondences from the internal distance structure alone and thereby tests fine-grained alignment. We further assess generalization on a curated test set of concepts non-overlapping with the training data. We show that fine-tuning pre-trained DNNs with Relational Knowledge Distillation (RKD), an established relational transfer method, brings DNNs close enough to humans to be aligned at the individual-object level on this test set. We also show that this improvement is driven by a more human-like global structure, as reflected in the ordering of distances among coarse categories, while the local human-DNN nearest-neighbor overlap rate remains largely unchanged. These findings indicate that relational transfer from humans brings the global structure of pre-trained DNNs close enough to the human structure to enable fine-grained human-DNN alignment without supervision.

Biological heterogeneity in Alzheimer's Disease (AD) poses a critical diagnostic challenge, particularly for traditional linear methods that fail to capture non-linear neural dynamics. To address this, we propose a diagnostic framework utilizing the Large Brain Model (LaBraM), pretrained on over 2,500 hours of EEG data. By integrating these high-dimensional latent embeddings with a non-linear Random Forest classifier, our approach effectively isolates robust disease markers. Under a rigorous subject-independent 5-fold cross-validation protocol, the method achieves an ROC-AUC of 89.36% +/- 3.49%, PR AUC of 81.45% +/- 4.43%, and Balanced Accuracy of 82.44% +/- 4.34% in distinguishing dementia patients from healthy controls. Notably, this performance uses only 8-second EEG segments, surpassing traditional spectral baselines, including band-power and parameterized oscillatory features (FOOOF). Post-hoc occlusion analysis confirms the model captures clinically validated biomarkers, specifically occipital-frontal Alpha and Theta rhythm degradation. Additional neurophysiological alignment analysis demonstrated that higher LaBraM-predicted dementia probability significantly correlated with worse cognitive performance, greater clinical severity, increased theta and alpha relative power, and higher aperiodic exponent. These findings demonstrate that deep latent representations extract clinically relevant signatures from noisy signals, enabling precise, rapid, and data-efficient diagnosis.

Recent research grounded in experiments has connected neural ring models of vision to how animals navigate a complex landscape of attractive targets. In this paper we investigate the mathematical and biological implications of a three-stage model where animals pre-process visual stimuli to identify a discrete set of targets, process this input to select the dominant targets, and then post-process this information to navigate the landscape. Incorporating finite target sizes and a neural density allows consistent target acquisition and pursuit reflecting what is seen in nature. Mathematically, we show this model corresponds to an energy minimization problem, simplifying both its analysis and numerical implementation. Biologically, we argue that the model is physiologically motivated, reproduces experimental observations, and presents a fairly direct pathway to decoding neural geometry via empirical data. Our results also demonstrate an analysis pipeline for future model extensions and suggest specific avenues for experimental research to test and parameterize our quantitative framework.