Data & Analytics

Analysis methodology for structural plasticity in neural networks — evaluating growth vs pruning operators, newborn unit integration stability, and time-sensitive optimization dynamics. Covers forward-active backward-starved phenomenon, insertion stability, and continual learning plasticity.

Multi-signal model of adult language learning using transformer brain alignment. Prediction shapes group-level neural architecture, feedback explains individual differences. fMRI-based with 102 subjects over 7 days. Activation: language learning, predictive coding, feedback signals, brain-model alignment, individual differences, transformer language models, artificial language learning, fMRI language representation.

LARGE: Locally Adaptive Regularization for estimating Gaussian Graphical Models — improving brain network connectivity estimation via node-specific penalty tuning. Activation: Gaussian graphical model, GGM, graphical Lasso, GLASSO, brain connectivity, functional connectivity, precision matrix, adaptive regularization, network neuroscience.

Robust volatility updates for Hierarchical Gaussian Filtering (HGF). Improves stability and convergence of uncertainty estimation in perceptual inference. Activation: hierarchical gaussian filter, volatility update, perceptual inference, active inference, uncertainty estimation.

早停策略技能 - 利用中间答案的置信度动态来决定何时终止推理，适用于大推理模型的长链式思维生成。基于论文 Early Stopping for Large Reasoning Models via Confidence Dynamics (arXiv 2604.04930)。激活关键词: 早停, early stop, confidence dynamics, reasoning stop, 推理终止, overthinking prevention, 防止过度思考。

Conservative adaptive rank methodology for quantum kinetic simulations — ACA SVD with Fermi-Dirac reconstruction preserving discrete macroscopic invariants near machine precision.

Separating wiring-specific from statistical control of dynamics in a complete connectome. Analysis of larval Drosophila brain showing coarse statistics set dynamical regime while specific wiring determines activity routing.

Connectome-Constrained Neural Network (CCNN) methodology for brain-inspired AI. Integrates biological structural connectivity (connectome) into artificial neural network architectures to improve generalization and biological plausibility. Activation: connectome constraint, structural connectivity, brain-inspired architecture, connectome-based AI, wiring cost, brain network prior, diffusion MRI connectivity.

Congestion-Aware Dynamic Axonal Delay mechanism for Spiking Neural Networks. Decomposes delay into channel-wise static base delay + global activity-conditioned shift. Reduces delay parameters by ~50% while improving accuracy on temporal tasks. Source: arXiv:2605.01291 (Bai et al., May 2026).

Congestion-Aware Dynamic Axonal Delay for Spiking Neural Networks. Replaces static per-synapse delays with input-dependent dynamic delays that adapt to network activity patterns, reducing delay parameters while improving temporal task performance. Activation: congestion-aware delay, dynamic axonal delay SNN, input-dependent delay, SNN temporal processing, adaptive delay learning.

早停策略技能 - 利用中间答案的置信度动态来决定何时终止推理，适用于大推理模型的长链式思维生成。基于论文 Early Stopping for Large Reasoning Models via Confidence Dynamics (arXiv 2604.04930)。激活关键词: 早停, early stop, confidence dynamics, reasoning stop, 推理终止, overthinking prevention, 防止过度思考。

Compositional quantum heuristics for mitigating barren plateaus in quantum machine learning. Assembles larger quantum models from smaller subcomponents with group-invariant loss functions introducing symmetry-induced inductive bias for improved gradient behavior. Use when: barren plateau mitigation, quantum graph neural networks, permutation-equivariant quantum models, recursive quantum-classical hybrid optimization, QIRO-inspired quantum heuristics, max-clique quantum detection, group-invari...

Hybrid quantum-classical machine learning using neutral-atom (cold-atom) reservoir computing for classification tasks, especially medical imaging. Covers the pipeline of guided auto-encoder dimensionality reduction, surrogate-driven training, and cold-atom reservoir state evolution. Use when: (1) implementing reservoir computing with quantum/neutral-atom systems, (2) building hybrid quantum-classical ML pipelines, (3) medical image classification with reservoir computing, (4) surrogate-gradie...

Coherence law for trainability in noisy equivariant quantum neural networks. U(1)-equivariant QNNs with light-cone gradient confinement, sector coherence rate as Rayleigh quotient, and open-system training law. Use when designing symmetric QNNs for noisy hardware, analyzing gradient survival under decoherence, or building noise-resilient quantum neural architectures.

Cognition-Inspired Dual-Stream Semantic Enhancement (DuSE) for Vision-Based Dynamic Emotion Modeling. Implements hierarchical temporal prompt clusters (HTPC) for cognitive priming and latent semantic emotion aggregators (LSEA) for knowledge integration. Models neuro-cognitive mechanisms from Conceptual Act Theory for dynamic facial expression recognition. Use for: emotion recognition, cognitive-inspired computer vision, neuro-cognitive modeling, dynamic facial expression analysis.

Coarse feedback for human-aligned visual representations. Use when: studying how supervisory signal granularity affects brain alignment in neural networks, designing brain-aligned vision models with minimal supervision, comparing coarse vs fine-grained training objectives, deriving coarse category labels from pretrained embeddings (PCA-based splits), representational similarity analysis (RSA) of neural/behavioral alignment, building AI systems aligned with human perception, or investigating w...

Hybrid CNN-SNN architecture for EEG-based imagined speech decoding. First integration of spiking neural networks into imagined speech BCI, achieving 80.13% accuracy on BCI Competition III benchmark. Activation: imagined speech, EEG decoding, CNN-SNN hybrid, spike-based BCI, neuromorphic BCI

Scalable neuromorphic computing via autonomous spiking dynamics in clockless (asynchronous) digital circuits implemented on FPGAs. Boolean spiking neurons with configurable excitatory/inhibitory weights, spike-encoded data processing pipeline. Bridges gap to analog neuromorphic systems without specialized hardware. Based on Oliveira Gomes & Rontani (arXiv: 2605.16114). Use when designing energy-efficient neuromorphic systems on FPGAs, exploring clockless asynchronous digital circuits for neur...

Circuit-level spiking neuron model for hardware robustness analysis. Studies how transistor-level variations affect SNN reliability on neuromorphic chips. Activation: circuit-level SNN, neuromorphic hardware reliability, transistor variation spiking, hardware spiking neuron, CMOS spiking, SNN fault tolerance

CFSPMNet - Cross-subject Fourier-guided Spatial-Patch Mamba Network for EEG Motor Imagery Decoding in Stroke Patients. Use when working with MI-EEG decoding, cross-subject BCI for stroke rehabilitation, Mamba-based EEG models, or Fourier-domain token reorganization for neural decoding.

Two-site cavity method for analyzing large nonlinear recurrent neural networks. Derives linear equivalence of nonlinear RNNs, computes full covariance matrices for specific quenched realizations, and separates Gaussian from non-Gaussian contributions in recurrent network dynamics. Use when analyzing: (1) high-dimensional RNN covariance structure, (2) nonlinear-to-linear network equivalence, (3) cavity method applications to neural dynamics, (4) quenched disorder in recurrent networks.

Convolutional Neural Network framework for detecting gaseous microemboli (GME) during cardiac procedures using transthoracic ultrasound. Activation triggers: emboli detection, cardiac ultrasound, microemboli GME, surgical safety, transcatheter monitoring

CaMBRAIN methodology for real-time continuous EEG inference using causal Mamba state space models. First model enabling long-range streaming inference of variable-length EEG signals with >10x higher throughput.

Calibrated measurement framework using the Brody exponent β as a quantitative measure of short-range exclusion in 2D spatial point processes. Originally from quantum chaos level-spacing statistics, now calibrated for spatial analysis with corrected CSR baseline, empirical β-r_excl calibration (Spearman ρ=0.988), and control protocols. Use for quantum chaos analysis, spatial statistics, prime number embeddings, and manufactured surface characterization.