Neural State Research Articles

Solving the Electronic Schrödinger Equation by Pairing Tensor-Network State with Neural Network Quantum State

Neural network methods have shown promise for solving complex quantum many-body systems. In this study, we develop a novel approach through incorporating the density-matrix renormalization group (DMRG) method with the neural network quantum state method. The results demonstrate that, when tensor-network pre-training is introduced into the neural network, a high efficiency can be achieved for quantum many-body systems with strong correlations.

Math can’t Move Matter

Abstract Causal platonism asserts that mathematical objects cause neural states in human brains. I raise the following four objections to it. (i) Quantum entanglement does not show that one object can causally affect another, although one is nontemporal, nonspatial, and unchanging. (ii) Causal platonism can neither be justified a posteriori nor a priori. (iii) To postulate mathematical media to flesh out mathematical causation is to multiply mysteries beyond necessity. (iv) To say that mathematical causation is unintelligible and inexplicable is not to complain about causation in general.

Neural Geometrodynamics, Complexity, and Plasticity: A Psychedelics Perspective.

We explore the intersection of neural dynamics and the effects of psychedelics in light of distinct timescales in a framework integrating concepts from dynamics, complexity, and plasticity. We call this framework neural geometrodynamics for its parallels with general relativity's description of the interplay of spacetime and matter. The geometry of trajectories within the dynamical landscape of "fast time" dynamics are shaped by the structure of a differential equation and its connectivity parameters, which themselves evolve over "slow time" driven by state-dependent and state-independent plasticity mechanisms. Finally, the adjustment of plasticity processes (metaplasticity) takes place in an "ultraslow" time scale. Psychedelics flatten the neural landscape, leading to heightened entropy and complexity of neural dynamics, as observed in neuroimaging and modeling studies linking increases in complexity with a disruption of functional integration. We highlight the relationship between criticality, the complexity of fast neural dynamics, and synaptic plasticity. Pathological, rigid, or "canalized" neural dynamics result in an ultrastable confined repertoire, allowing slower plastic changes to consolidate them further. However, under the influence of psychedelics, the destabilizing emergence of complex dynamics leads to a more fluid and adaptable neural state in a process that is amplified by the plasticity-enhancing effects of psychedelics. This shift manifests as an acute systemic increase of disorder and a possibly longer-lasting increase in complexity affecting both short-term dynamics and long-term plastic processes. Our framework offers a holistic perspective on the acute effects of these substances and their potential long-term impacts on neural structure and function.

Differential Formation of Motor Cortical Dynamics during Movement Preparation According to the Predictability of Go Timing.

The motor cortex not only executes but also prepares movement, as motor cortical neurons exhibit preparatory activity that predicts upcoming movements. In movement preparation, animals adopt different strategies in response to uncertainties existing in nature such as the unknown timing of when a predator will attack-an environmental cue informing "go." However, how motor cortical neurons cope with such uncertainties is less understood. In this study, we aim to investigate whether and how preparatory activity is altered depending on the predictability of "go" timing. We analyze firing activities of the anterior lateral motor cortex in male mice during two auditory delayed-response tasks each with predictable or unpredictable go timing. When go timing is unpredictable, preparatory activities immediately reach and stay in a neural state capable of producing movement anytime to a sudden go cue. When go timing is predictable, preparation activity reaches the movement-producible state more gradually, to secure more accurate decisions. Surprisingly, this preparation process entails a longer reaction time. We find that as preparatory activity increases in accuracy, it takes longer for a neural state to transition from the end of preparation to the start of movement. Our results suggest that the motor cortex fine-tunes preparatory activity for more accurate movement using the predictability of go timing.

Application of Gradient Optimization Methods in Defining Neural Dynamics

Applications of gradient method for nonlinear optimization in development of Gradient Neural Network (GNN) and Zhang Neural Network (ZNN) are investigated. Particularly, the solution of the matrix equation AXB=D which changes over time is studied using the novel GNN model, termed as GGNN(A,B,D). The GGNN model is developed applying GNN dynamics on the gradient of the error matrix used in the development of the GNN model. The convergence analysis shows that the neural state matrix of the GGNN(A,B,D) design converges asymptotically to the solution of the matrix equation AXB=D, for any initial state matrix. It is also shown that the convergence result is the least square solution which is defined depending on the selected initial matrix. A hybridization of GGNN with analogous modification GZNN of the ZNN dynamics is considered. The Simulink implementation of presented GGNN models is carried out on the set of real matrices.

Integrating Large Language Model, EEG, and Eye-Tracking for Word-Level Neural State Classification in Reading Comprehension.

With the recent proliferation of large language models (LLMs), such as Generative Pre-trained Transformers (GPT), there has been a significant shift in exploring human and machine comprehension of semantic language meaning. This shift calls for interdisciplinary research that bridges cognitive science and natural language processing (NLP). This pilot study aims to provide insights into individuals' neural states during a semantic inference reading-comprehension task. We propose jointly analyzing LLMs, eye-gaze, and electroencephalographic (EEG) data to study how the brain processes words with varying degrees of relevance to a keyword during reading. We also use feature engineering to improve the fixation-related EEG data classification while participants read words with high versus low relevance to the keyword. The best validation accuracy in this word-level classification is over 60% across 12 subjects. Words highly relevant to the inference keyword received significantly more eye fixations per word: 1.0584 compared to 0.6576, including words with no fixations. This study represents the first attempt to classify brain states at a word level using LLM-generated labels. It provides valuable insights into human cognitive abilities and Artificial General Intelligence (AGI), and offers guidance for developing potential reading-assisted technologies.

Optimally Tuned Gated Recurrent Unit Neural Network-Based State of Health Estimation Scheme for Lithium Ion Batteries

Optimally Tuned Gated Recurrent Unit Neural Network-Based State of Health Estimation Scheme for Lithium Ion Batteries

Design and Evaluation of Low Voltage Neural Network-Based State Estimators in Scenarios With Minimal Measurement Infrastructure

Design and Evaluation of Low Voltage Neural Network-Based State Estimators in Scenarios With Minimal Measurement Infrastructure

Single-nucleus expression characterization of non-enhancing region of recurrent high-grade glioma.

Non-enhancing (NE) infiltrating tumor cells beyond the contrast-enhancing (CE) bulk of tumor are potential propagators of recurrence after gross total resection of high-grade glioma. We leveraged single-nucleus RNA sequencing on 15 specimens from recurrent high-grade gliomas (n = 5) to compare prospectively identified biopsy specimens acquired from CE and NE regions. Additionally, 24 CE and 22 NE biopsies had immunohistochemical staining to validate RNA findings. Tumor cells in NE regions are enriched in neural progenitor cell-like cellular states, while CE regions are enriched in mesenchymal-like states. NE glioma cells have similar proportions of proliferative and putative glioma stem cells relative to CE regions, without significant differences in % Ki-67 staining. Tumor cells in NE regions exhibit upregulation of genes previously associated with lower grade gliomas. Our findings in recurrent GBM paralleled some of the findings in a re-analysis of a dataset from primary GBM. Cell-, gene-, and pathway-level analyses of the tumor microenvironment in the NE region reveal relative downregulation of tumor-mediated neovascularization and cell-mediated immune response, but increased glioma-to-nonpathological cell interactions. This comprehensive analysis illustrates differing tumor and nontumor landscapes of CE and NE regions in high-grade gliomas, highlighting the NE region as an area harboring likely initiators of recurrence in a pro-tumor microenvironment and identifying possible targets for future design of NE-specific adjuvant therapy. These findings also support the aggressive approach to resection of tumor-bearing NE regions.

Development and Evaluation of a Real-Time Phase-Triggered Stimulation Algorithm for the CorTec Brain Interchange.

With the development and characterization of biomarkers that may reflect neural network state as well as a patient's clinical deficits, there is growing interest in more complex stimulation designs. While current implantable neuromodulation systems offer pathways to expand the design and application of adaptive stimulation paradigms, technological drawbacks of these systems limit adaptive neuromodulation exploration. In this paper, we discuss the implementation of a phase-triggered stimulation paradigm using a research platform composed of an investigational system known as the CorTec Brain Interchange (CorTec GmbH, Freiburg, Germany), and an open-source software tool known as OMNI-BIC. We then evaluate the stimulation paradigm's performance in both benchtop and in vivo human demonstrations. Our findings indicate that the Brain Interchange and OMNI-BIC platform is capable of reliable administration of phase-triggered stimulation and has the potential to help expand investigation within the adaptive neuromodulation design space.

Clinical and neurophysiological manifestations of sluggish cognitive tempo in children

To study the clinical and neurophysiological features of children with low cognitive tempo (NCT), as well as the effectiveness of the drug Pantogam in the treatment of this pathology. A total of 90 children aged 8 to 10 years were examined. Of these, the main study group consisted of 30 children with NCT, the comparison group consisted of 30 children with a combined type of attention deficit hyperactivity disorder ADHD (ADHD-K), the control group consisted of 30 children without neuropsychiatric disorders. The study used clinical, neurophysiological (electroencephalography (EEG)) and parametric methods. The CMAS scale of apparent anxiety (The Children's Form of Manifest Anxiety Scale), the SNAP-IY scale (assessment of the degree of inattention, hyperactivity and impulsivity), the TOVA computer test (the Test of Variables of Attention), the scale «SCT» (Sluggish Cognitive Tempo) for assessing manifestations of low cognitive tempo, the «RAM» technique for quantifying working memory. Pantogam was used to treat patients at a dose of 750 mg per day for 8 weeks. Patients with NCT are characterized by more pronounced attention disorders compared with healthy peers and with children with ADHD-K, and they have a decrease in mainly not selective attention, but the overall level of functional activity. Also, the group of children with NCT has an increased level of anxiety compared to the group of children with ADHD. A comparative analysis of the level of impulsivity showed that children with NCT are less characterized by a deficit in inhibition processes. According to the quantitative analysis of the EEG, specific changes in functional activity in the frontal and central regions of the cerebral cortex were revealed (a statistically significant increase in the ratio of absolute theta rhythm to beta1 rhythm, compared with other groups), reflecting insufficient cortical arousal and less focused neural states. When re-evaluating the condition of children with NCT after a course of therapy with Pantogam, an improvement in the form of a decrease in the degree of inattention, the severity of memory impairment and a decrease in reaction time was recorded in 60% of cases. According to quantitative EEG analysis, there was a significant decrease in the ratio of absolute theta rhythm to beta1 rhythm in the central leads of both hemispheres and in the parietal-temporal leads of the left hemisphere, indicating an increase in the level of overall activation of the cerebral cortex after a course of treatment. Clinical and neurophysiological differences were revealed in patients with NCT and with combined ADHD. It has been shown that the use of Pantogam for the treatment of children with NCT leads not only to a decrease in the main manifestations of this disorder, but also to an improvement in the functional state of the brain.

Multi-Action Knee Contact Force Prediction by Domain Adaptation.

Most recent musculoskeletal dynamics estimation methods are designed for predefined actions, such as gait, and don't generalize to various tasks. In this work, we address the problem of estimating internal biomechanical forces during more than one actions by introducing unsupervised domain adaptation into a deep learning model. More specifically, we developed a Bidirectional Long Short-Term Memory network for knee contact force prediction, enhanced with correlation alignment layers, in order to minimize the domain shift between kinematic data from different actions. Furthermore, we used the novel Neural State Machine (NSM) as a simulation platform to test and visualize our model predictions in a wide range of trajectories adapted to different 3D scene geometries in real-time. We conducted multiple experiments, including comparison with previous models, model alignment across action classes and real-to-synthetic data alignment. The results showed that the proposed deep learning architecture with domain adaptation performs better than the benchmark in terms of NRMSE and t-test. Overall, our method is capable of predicting knee contact forces for more than one action classes using a single architecture and thereby opens the path for estimating internal forces for intermediate actions, while the knowledge of the hidden state of motion may be used to support personalized rehabilitation. Moreover, our model can be easily integrated into any human motion simulation environment, which shows its potential in enabling biomechanical analysis in an automated and computationally efficient way.

Online Estimating Pairwise Neuronal Functional Connectivity in Brain-Machine Interface.

Neurons respond to external stimuli and form functional networks through pairwise interactions. A neural encoding model can describe a single neuron's behavior, and brain-machine interfaces (BMIs) provide a platform to investigate how neurons adapt, functionally connect, and encode movement. Movement modulation and pairwise functional connectivity are modeled as high-dimensional tuning states, estimated from neural spike train observations. However, accurate estimation of this neural state vector can be challenging as pairwise neural interactions are highly dimensional, change in different temporal scales from movement, and could be non-stationary. We propose an Adam-based gradient descent method to online estimate high-dimensional pairwise neuronal functional connectivity and single neuronal tuning adaptation simultaneously. By minimizing negative log-likelihood based on point process observation, the proposed method adaptively adjusts the learning rate for each dimension of the neural state vectors by employing momentum and regularizer. We test the method on real recordings of two rats performing the brain control mode of a two-lever discrimination task. Our results show that our method outperforms existing methods, especially when the state is sparse. Our method is more stable and faster for an online scenario regardless of the parameter initializations. Our method provides a promising tool to track and build the time-variant functional neural connectivity, which dynamically forms the functional network and results in better brain control.

Classifying Routine Clinical Electroencephalograms With Multivariate Iterative Filtering and Convolutional Neural Networks.

Electroencephalogram (EEG) is widely used in basic and clinical neuroscience to explore neural states in various populations, and classifying these EEG recordings is a fundamental challenge. While machine learning shows promising results in classifying long multivariate time series, optimal prediction models and feature extraction methods for EEG classification remain elusive. Our study addressed the problem of EEG classification under the framework of brain age prediction, applying a deep learning model on EEG time series. We hypothesized that decomposing EEG signals into oscillatory modes would yield more accurate age predictions than using raw or canonically frequency-filtered EEG. Specifically, we employed multivariate intrinsic mode functions (MIMFs), an empirical mode decomposition (EMD) variant based on multivariate iterative filtering (MIF), with a convolutional neural network (CNN) model. Testing a large dataset of routine clinical EEG scans (n = 6540) from patients aged 1 to 103 years, we found that an ad-hoc CNN model without fine-tuning could reasonably predict brain age from EEGs. Crucially, MIMF decomposition significantly improved performance compared to canonical brain rhythms (from delta to lower gamma oscillations). Our approach achieved a mean absolute error (MAE) of 13.76 ± 0.33 and a correlation coefficient of 0.64 ± 0.01 in brain age prediction over the entire lifespan. Our findings indicate that CNN models applied to EEGs, preserving their original temporal structure, remains a promising framework for EEG classification, wherein the adaptive signal decompositions such as the MIF can enhance CNN models' performance in this task.

Fixed-time command filtered output feedback control for twin-roll inclined casting system with prescribed performance.

The article investigates the issue of fixed-time control with adaptive output feedback for a twin-roll inclined casting system (TRICS) with disturbance. First, by using the mean value theorem, the nonaffine functions are decoupled to simplify the system. Second, radial basis function neural networks (RBFNNs) are introduced to approximate an unknown term, and a nonlinear neural state observer is created to handle the effects of unmeasured states. Then, the backstepping design framework is combined with prescribed performance and command filtering techniques to demonstrate that the scheme proposed in this article guarantees system performance within a fixed-time. The control design parameters determine the upper bound of settling time, regardless of the initial state of the system. Meanwhile, it ensures that all signals in the closed-loop system (CLS) remain bounded, and it can also maintain the tracking error within a predefined range within a fixed time. Finally, simulation results assert the effectiveness of the method.

Disturbance estimation for the hydraulic drive system based on the state observer and neural network

The hydraulic servo drive system has the ability to operate stably with high reliability in conditions of continuous operation and large capacity. Aside from these advantages, the system always has nonlinear and uncertain factors that affect and reduce the quality of the controller. These factors include the affected disturbances and the uncertainty of the system model. This paper proposes a method to identify the sum of nonlinear effects and disturbances for the hydraulic drive system, which uses the axial piston pump, based on the application of the Radial Basic Functions neural network (RBF) and state observer model. The simulation results show that the total disturbances and nonlinear effects can be accurately estimated in real time through the weight updating rule.

Neural network quantum states analysis of the Shastry-Sutherland model

We utilize neural network quantum states (NQS) to investigate the ground state properties of the Heisenberg model on a Shastry-Sutherland lattice using the variational Monte Carlo method. We show that already relatively simple NQSs can be used to approximate the ground state of this model in its different phases and regimes. We first compare several types of NQSs with each other on small lattices and benchmark their variational energies against the exact diagonalization results. We argue that when precision, generality, and computational costs are taken into account, a good choice for addressing larger systems is a shallow restricted Boltzmann machine NQS. We then show that such NQS can describe the main phases of the model in zero magnetic field. Moreover, NQS based on a restricted Boltzmann machine correctly describes the intriguing plateaus forming in magnetization of the model as a function of increasing magnetic field.

NEURAL STATES ASSOCIATED WITH MOTIVATION AND REWARD MODULATE AGING IN DROSOPHILA

Abstract Humans, scientists especially, like to believe that we are in control of our behaviors and that logical, rational thought is the driving force behind them. Even for the most conscious of actions, however, there are serious questions about whether this is true. Economists, for example, are well-versed in the notion that consumer behavior is irrational. Why? Psychologists attribute it to the complex evaluation of internal and external cues, which is often need-based but can be strongly influenced by perceived costs and benefits. Molecular neuroscience has gone a step further in recent years, often using simple model organisms, to provide a well-defined framework for dissecting the causes and consequences of motivated behaviors. Homeostatic needs are as influential in this process as are rewarding experiences, and I will discuss evidence that the neurons, neural circuits, and conserved signaling molecules that influence neural states associated with feeding, mating, danger in flies are also important modulators of healthy aging.

Implicit neural state functions in hybrid reliability analysis of plane frame

Implicit neural state functions in hybrid reliability analysis of plane frame

A three-dimensional model of neural activity and phenomenal-behavioral patterns.

How phenomenal experience and behavior are related to neural activity in physiology and psychopathology represents a fundamental question in neuroscience and psychiatry. The phenomenal-behavior patterns may be deconstructed into basic dimensions, i.e., psychomotricity, affectivity, and thought, which might have distinct neural correlates. This work provides a data overview on the relationship of these phenomenal-behavioral dimensions with brain activity across physiological and pathological conditions (including major depressive disorder, bipolar disorder, schizophrenia, attention-deficit/hyperactivity disorder, anxiety disorders, addictive disorders, Parkinson's disease, Tourette syndrome, Alzheimer's disease, and frontotemporal dementia). Accordingly, we propose a three-dimensional model of neural activity and phenomenal-behavioral patterns. In this model, neural activity is organized into distinct units in accordance with connectivity patterns and related input/output processing, manifesting in the different phenomenal-behavioral dimensions. (1) An external neuralunit, which involves the sensorimotor circuit/brain's sensorimotor network and is connected with the external environment, processes external inputs/outputs, manifesting in the psychomotor dimension (processing of exteroception/somatomotor activity). External unit hyperactivity manifests in psychomotor excitation (hyperactivity/hyperkinesia/catatonia), while external unit hypoactivity manifests in psychomotor inhibition (retardation/hypokinesia/catatonia). (2) An internal neuralunit, which involves the interoceptive-autonomic circuit/brain's salience network and is connected with the internal/body environment, processes internal inputs/outputs, manifesting in the affective dimension (processing of interoception/autonomic activity). Internal unit hyperactivity manifests in affective excitation (anxiety/dysphoria-euphoria/panic), while internal unit hypoactivity manifests in affective inhibition (anhedonia/apathy/depersonalization). (3) An associative neuralunit, which involves the brain's associative areas/default-mode network and is connected with the external/internal units (but not with the environment), processes associative inputs/outputs, manifesting in the thought dimension (processing of ideas). Associative unit hyperactivity manifests in thought excitation (mind-wandering/repetitive thinking/psychosis), while associative unit hypoactivity manifests in thought inhibition (inattention/cognitive deficit/consciousness loss). Finally, these neural units interplay and dynamically combine into various neural states, resulting in the complex phenomenal experience and behavior across physiology and neuropsychiatric disorders.