All authors
Claude Skills by hiyenwong
github.com/hiyenwong9,934 skills5 installs19,223 views
- Contextual Quantum Neural Stock PredictionContextual quantum neural network methodology for multi-asset stock price prediction using quantum batch gradient update (QBGU) and quantum multi-task learning (QMTL) with share-and-specify ansatz.Votes: 0GitHub stars: 3
- Contextual Role Object GeometryContextual Role Modulates Object Representational Geometry in the Human Brain. fMRI study showing how object representations are dynamically remapped based on contextual role (action target vs passive element), with double dissociation between action affordance and semantic representational organization. Activation: representational geometry, fMRI object recognition, contextual modulation, action affordance, ventral dorsal stream, brain network remapping, naturalistic neuroscience, object rep...Votes: 0GitHub stars: 3
- Contextual Role Object Representational Geometry**arXiv:2605.23111** | Submitted: 22 May 2026 | q-bio.NCVotes: 0GitHub stars: 3
- Continual Learning Spiking TransformerCATFormer: When Continual Learning Meets Spiking Transformers With Dynamic Thres - Spiking transformer architecture for continual learning. Activation triggers: continual, learning, spiking, neuroscience, SNN.Votes: 0GitHub stars: 3
- Attribution Neuron UtilityGradient-based eXplained Deviation (GXD) neuron utility metric for plasticity restoration in deep networks. Combines attribution-guided neuron importance scoring with critical bottleneck period (CBP) detection and targeted reset. Use when: catastrophic forgetting, plasticity loss, neuron-level analysis, continual learning reset methods, attribution-guided training.Votes: 0GitHub stars: 3
- Characterizing Target ShiftOnline kernel regression effective target shift theory and correction. Proves online learning is equivalent to offline learning with shifted targets, derives label correction to achieve offline-optimal performance. Use when: online learning, continual learning with distribution shift, target shift correction, kernel regression analysis, EWA equivalence.Votes: 0GitHub stars: 3
- ComemnetCoMemNet: Contrastive sampling with Memory Replay Network for continual traffic prediction. Dual-branch architecture with contrastive positive/negative sample generation for efficient memory replay in spatio-temporal forecasting. Use when: traffic prediction, spatio-temporal forecasting, continual learning with memory replay, contrastive CL.Votes: 0GitHub stars: 3
- Craft ClCRAFT: Forgetting-Aware Intervention-Based Adaptation for continual learning. Avoids weight updates by learning low-rank interventions on hidden representations. Routes in representation space using KL divergence to decide between adaptation and routing. Use when: LLM continual learning, intervention-based adaptation, catastrophic forgetting mitigation, representation-space routing.Votes: 0GitHub stars: 3
- Scene Adaptive MoeScene-Adaptive Mixture of Experts (SAMoE-C) for continual learning in CSI-based human activity recognition. Uses MoE architecture with domain-specific expert routing and scene-adaptive gating for domain-incremental HAR. Use when: CSI sensing, HAR, domain-incremental learning, MoE continual learning, wireless sensing CL.Votes: 0GitHub stars: 3
- Continual Robot Policy Variational Neural DynamicsContinual robot policy learning framework using Variational Neural Dynamics. Combines analytical physics prior with neural residual for unmodeled effects. Recurrent encoder infers hidden conditions from recent interaction. Policy learning via differentiable simulation with sampled dynamics. Deployment uses online condition inference for recurring dynamics recovery. Activation: continual robot learning, variational dynamics, hidden condition, physics prior, differentiable simulation, quadrotor...Votes: 0GitHub stars: 3
- Continuous Metadata Conditioning PeftContinuous (non-discretized) metadata conditioning for parameter-efficient VL/CLIP adaptation — feed numerical attributes directly into the prompt representation so the embedding space modulates smoothly, while inference stays purely visual (no metadata needed at test). Use when adapting vision-language models to longitudinal/temporal distribution shift where discretizing metadata into text loses signal.Votes: 0GitHub stars: 3
- Contrastive On Policy ThinkingCopT methodology for LLM reasoning - answer-first thinking with continuous embedding contrastive verifiers and dynamic KL-based reliability estimation for efficient agentic reasoningVotes: 0GitHub stars: 3
- Contrastive Semantic Projection Neuron LabelingContrastive Semantic Projection (CSP) for faithful neuron labeling in deep networks using contrastive examples. Two-stage pipeline with VLM-based candidate generation and CLIP-based label assignment. Improves interpretability and explanation faithfulness. Activation: neuron labeling, contrastive examples, neural network interpretability, feature visualization, semantic projection.Votes: 0GitHub stars: 3
- Contravariance Theory Strong AlignmentContravariance Theory methodology — formal proof that minimal DNN solutions to sufficiently hard tasks exhibit strong alignment: weak alignment of representations guarantees strong alignment of privileged axes, and alignment zippers up the network hierarchy, proving convergent evolution is inevitable.Votes: 0GitHub stars: 3
- Coarse Feedback Visual AlignmentCoarse feedback for human-aligned visual representations. Demonstrates that extremely coarse classification signals (e.g., 8 classes) produce representations that match or exceed brain alignment of fine-grained (1000-class) or self-supervised models. Use when: studying visual-brain alignment, computational neuroscience, brain-inspired vision models, training signal granularity, representational similarity analysis.Votes: 0GitHub stars: 3
- Early Stopping Confidence DynamicsImplement early stopping strategies for reasoning models based on confidence dynamics. Detects when reasoning trajectories reach high confidence early vs. unproductive long reasoning traces. Use for: (1) optimizing LLM reasoning tasks, (2) reducing computational cost, (3) preventing overthinking, (4) improving accuracy-compute tradeoff. Based on research: arXiv:2604.04930 'Early Stopping for Large Reasoning Models via Confidence Dynamics'Votes: 0GitHub stars: 3
- Fade Adaptive Weight DecayFADE: Forgetting through Adaptive Decay for continual learning. Adapts per-parameter weight decay rates online via meta-gradient descent. Balances acquiring new knowledge with retaining old. Activation: FADE, adaptive weight decay, continual learning forgetting, meta-gradient weight decay, controlled forgetting, Ramesh Schmidhuber.Votes: 0GitHub stars: 3
- Fast Efficient Coding Gain AdaptiveFast efficient coding and sensory adaptation in gain-adaptive recurrent networks — unified mechanistic model reconciling adapter-repulsion and prior-attraction phenomena via gain modulation.Votes: 0GitHub stars: 3
- Feedforward Dynamics Stimulus EncodingThe illusory simplicity of the feedforward pass — evidence for dynamical nature of stimulus encoding in neural networks. Demonstrates that feedforward computation involves complex temporal dynamics rather than static transformations. Applicable to neural network analysis, computational neuroscience, dynamical systems analysis. 触发词: feedforward dynamics, stimulus encoding, temporal computation, dynamical neural networks, illusory simplicityVotes: 0GitHub stars: 3
- Flexible Phase Locking Cortical ThetaDynamical systems methodology for flexible phase-locking in cortical oscillators. Multi-timescale inhibitory currents enable entrainment to rhythms slower than intrinsic frequency via delayed Hopf bifurcation. Activation: phase-locking, cortical oscillators, theta oscillations, speech segmentation, delayed Hopf, multi-timescale dynamics, entrainment, inhibitory currents.Votes: 0GitHub stars: 3
- Geodynamics Geometric State SpaceGeometric State-Space Neural Network for brain dynamics modeling. Combines state-space models with geometric constraints on brain connectivity to capture latent neural state evolution. Applicable to fMRI/EEG dynamics modeling, functional neuroimaging analysis, and brain network temporal dynamics. Trigger: state-space models brain dynamics, geometric neural networks, fMRI dynamics, latent neural states, brain connectivity geometryVotes: 0GitHub stars: 3
- Hgf Robust Volatility UpdatesRobust 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.Votes: 0GitHub stars: 3
- Koopman Stability Preserving IdStability-preserving system identification using Koopman operator lifting with ISS-LMI constraints. Enables data-driven modeling of nonlinear systems (especially Persidskii-class and electromechanical systems) while guaranteeing input-to-state stability. Use when: (1) identifying nonlinear system models from trajectory data, (2) needing stability guarantees in learned models, (3) working with Persidskii systems or sector-bounded nonlinearities, (4) designing robust observers for partially-mea...Votes: 0GitHub stars: 3
- Kuramoto Oscillatory Phase EncodingKuramoto Oscillatory Phase Encoding for Vision Transformers - neuro-inspired synchronization-based phase encoding that mimics biological oscillatory neural dynamics. Uses Kuramoto model to encode spatial information through phase relationships for efficient vision transformers. Activation: kuramoto phase encoding, oscillatory encoding, vision transformer phase, biological synchronization, neural oscillator encoding.Votes: 0GitHub stars: 3
- Large Adaptive Regularization Gaussian GraphicalLARGE: 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.Votes: 0GitHub stars: 3
- Narx Topological Phase MappingNARX neural network methodology for deterministic mapping of topological phase transitions in quantum systems. Uses autoregressive exogenous inputs to discover functional identities between topological invariants and critical parameters.Votes: 0GitHub stars: 3
- Neurobridge Koopman DynamicsNeuroBRIDGE: Behavior-conditioned Koopman dynamics with Riemannian alignment for brain network analysis. Uses Koopman operator theory with Riemannian geometry for dynamic functional connectivity modeling to predict substance use initiation risk. Activation: Koopman brain dynamics, behavior-conditioned Koopman, Riemannian alignment, dynamic connectivity, SUI prediction, adolescent brain.Votes: 0GitHub stars: 3
- Neuromechanical Locomotion DynamicsNeuromechanical modeling framework that connects neural activity to behavioral locomotion dynamics. Combines spectral mode representations with Helmholtz-Nambu decompositions and Bayesian inference to infer predictive stochastic models from neural population data. Activation: neuromechanics, locomotion dynamics, neural-behavior mapping, Helmholtz-Nambu, C. elegans, optogenetic control, behavior prediction from neural activity.Votes: 0GitHub stars: 3
- Neuromorphic Oscillator Reservoir ComputingReservoir computing using parametrically-driven oscillators and frequency combs for neuromorphic computation. Three-regime system (sub-threshold, parametric resonance, frequency-comb) with 2:1 resonance. Optimal performance at parametric resonance boundary. Applications: chaotic time-series prediction (Mackey-Glass, Rössler, Lorenz), edge AI, analog neural networks.Votes: 0GitHub stars: 3
- Neurosymbolic Robustness AnalysisNeurosymbolic robustness analysis framework for discrete systems. Uses LLM neural reasoning layer + symbolic verification layer to analyze robustness of discrete-event systems against transition deviations. Based on arXiv:2606.03872 (Jun 2026).Votes: 0GitHub stars: 3
- Noise Induced Group Level Synchronization OscillatorsCommon noise-induced synchronization methodology for uncoupled oscillator groups. Demonstrates that groups receiving the same common noise synchronize at the collective level without inter-group coupling.Votes: 0GitHub stars: 3
- Ode Complexity DynamicsComplexity theory analysis of Ordinary Differential Equations. Examines computational complexity of ODE solutions, existence conditions, and complexity barriers. Trigger words: ODE complexity, computational complexity differential equations, ODE existence conditions, complexity barriers, numerical analysis complexity.Votes: 0GitHub stars: 3
- Parametric Oscillator Reservoir ComputingNeuromorphic reservoir computing using parametrically-driven oscillators and frequency combs. Covers oscillator-based RC, 2:1 parametric resonance regimes, and bifurcation-driven computational capability mapping.Votes: 0GitHub stars: 3
- Parametrically Driven Oscillator NeuromorphicReservoir computing using parametrically-driven oscillators and frequency combs (arXiv:2604.21861). Demonstrates neuromorphic computation via two-mode parametric oscillator with 2:1 resonance across sub-threshold, parametric resonance, and frequency-comb regimes. Covers drive amplitude input encoding, temporal/spectral response sampling, chaotic time-series prediction (Mackey-Glass, Rossler, Lorenz), and design principles for tuning physical oscillator-based reservoir computers.Votes: 0GitHub stars: 3
- Predictive Feedback Signals Language RepresentationsMulti-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.Votes: 0GitHub stars: 3
- Qshift Adaptive SamplingqSHIFT adaptive sampling protocol for higher-order quantum simulation. Implements L-independent gate complexity with O(t^(1+r)) error scaling through adaptive distribution updates. Use when: performing quantum Hamiltonian simulation, designing quantum algorithms requiring high-precision time evolution, implementing qDRIFT-style randomized product formulas, or optimizing circuit depth for NISQ devices. Triggers on: qshift, adaptive sampling quantum, quantum simulation protocol, higher-order Tr...Votes: 0GitHub stars: 3
- Stability Goal ObfuscationStability-goal obfuscation tradeoff methodology for autonomous agents. Addresses the problem that Lyapunov-stable goal-directed trajectories are inherently legible to Bayesian observers, leaking intent. Combines control Lyapunov functions (CLFs), probabilistic control barrier functions (PCBFs), and Rao-Blackwellized particle filter (RBPF) belief-state analysis to maintain task stability while obfuscating intent from passive observers. Use when: designing privacy-preserving autonomous systems,...Votes: 0GitHub stars: 3
- State Adaptive Error CorrectionState-adaptive error correction and fault tolerance methodology. Use when designing resilient systems that need to adapt error handling based on current system state, optimize recovery strategies dynamically, or build fault-tolerant architectures. Applicable to distributed systems, quantum computing, network protocols, ML pipelines, and control systems. Trigger words: state-adaptive, adaptive error correction, fault-tolerant design, resilient architecture, dynamic error handling, system state...Votes: 0GitHub stars: 3
- Stp Stabilizes Goal Conditioned Dynamics短时程突触可塑性(STP)稳定目标条件化动力学方法论。研究STP如何在PFC储水池模型中支持多步目标导向行动规划，通过动态调节有效连接保持目标信息。Activation: STP, goal-conditioned dynamics, PFC reservoir, action planning, 突触可塑性, 目标导向行为.Votes: 0GitHub stars: 3
- Structural Plasticity Growth StabilityAnalysis 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.Votes: 0GitHub stars: 3
- Stuart Landau Oscillatory GnnComplex-Valued Stuart-Landau Graph Neural Network (SLGNN) — oscillatory GNN grounded in Stuart-Landau oscillator dynamics near Hopf bifurcations. Retains both amplitude and phase dynamics for rich phenomena like amplitude regulation and multistable synchronization. Activation: Stuart-Landau GNN, SLGNN, oscillatory graph neural network, Hopf bifurcation GNN, amplitude-phase GNN.Votes: 0GitHub stars: 3
- Syndrome Adaptive Gain QldpcSyndrome Adaptive Gain Min-Sum (SAGMS) decoding methodology for quantum LDPC codes. Dynamically adjusts the MS scaling factor based on syndrome patterns during iterative decoding, eliminating the need for per-code/per-noise-level gain optimization. Use when: designing quantum LDPC decoders, optimizing belief propagation alternatives, implementing low-complexity QEC decoding, comparing MS vs BP vs neural decoders, or working with generalized bicycle QLDPC codes. Keywords: quantum LDPC decoding...Votes: 0GitHub stars: 3
- Tide Ei DynamicsTIDE (Temporal Inhibitory-Excitatory Dynamic Engine) methodology — neuro-inspired architecture using asymmetric Excitatory-Inhibitory (E-I) networks with Wilson-Cowan dynamics and lateral inhibition for stabilized neural dynamics. Integrates Dale's principle (80:20 E-I ratio), hierarchical receptive fields, and game-theoretic energy-based optimization. Use when: designing neuro-inspired architectures with stability guarantees, building continuous thought/reasoning systems with internal dynami...Votes: 0GitHub stars: 3
- Tsodyks Markram Chaotic DynamicsTsodyks-Markram短时程突触可塑性的混沌动力学。研究确定性TM模型中Shilnikov同宿分岔导致混沌行为的路径，揭示网络动力学不可预测性和对初始条件的敏感性。适用于计算神经科学、突触可塑性建模、混沌动力学分析。触发词：短时程突触可塑性、Tsodyks-Markram模型、Shilnikov分岔、混沌动力学、short-term synaptic plasticity、Tsodyks-Markram model、Shilnikov homoclinic bifurcation、chaotic dynamics。Votes: 0GitHub stars: 3
- Tunneling Phase Diagram MlTunneling phase diagram methodology — machine learning framework for decoupling true quantum tunneling strength from composite kinetic isotope effects. For quantum chemistry, ML-driven quantum analysis, and kinetic modeling.Votes: 0GitHub stars: 3
- Ultrastructure To Dynamics CompilerSystematic methodology for compiling molecular ultrastructure into neural dynamics - bridging microscopic brain structure to computational function. Activation: ultrastructure compiler, molecular neural dynamics, connectome to function, structural biology, neural compilation.Votes: 0GitHub stars: 3
- Conv Delay Learning SnnCombining convolution and delay learning in recurrent spiking neural networks. Methodology for joint learning of synaptic weights and synaptic delays using modified STDP for enhanced spatiotemporal pattern recognition. Keywords: convolutional SNN, delay learning, spatiotemporal patterns, STDP, recurrent SNN, temporal coding.Votes: 0GitHub stars: 3
- Convergent Evolution Algorithmic SpaceFramework for analyzing convergent evolution in neural network weight structures during training. Uses matching-based comparison with permutation-invariant features and Hungarian matching to align hidden neurons, then applies structural distance metrics to identify task-specific attractors in weight space.Votes: 0GitHub stars: 3
- Convergent Evolution Neural Representation SpaceDBNs spontaneously organize representations by class without supervision.Votes: 0GitHub stars: 3
- Convex TokenizationConvexTok methodology — formulating tokenizer construction as a convex optimization (linear program) instead of greedy BPE/Unigram. Use when: (1) Designing tokenizers for new languages or domains, (2) Improving bits-per-byte (BpB) efficiency of LLM tokenizers, (3) Evaluating tokenizer quality beyond greedy heuristics, (4) Tokenizer research comparing BPE/Unigram vs. globally optimal approaches.Votes: 0GitHub stars: 3