メニュー
このリポジトリの skills
SkillsMP は hiyenwong/ai_collection から 4,114 件の skill を収集しています。skill を開くとソースと詳細を確認できます。
hiyenwong/ai_collection収集済み skill 4,114 件中 40 件を表示しています。
Analog Kolmogorov-Arnold Networks (AKANs) for low-power function approximation in flexible electronics. Hardware-software co-optimization with circuit-level error modeling during training and dual-level pruning. Targets wearable biosignal processing (EEG,…
原文の言語: 英語
模拟神经形态硬件上的多时间尺度突触可塑性方法。基于钙离子动力学的突触可塑性规则，结合突触标记捕获假说，在BrainScaleS-2系统上实现。适用于脉冲神经网络模拟、神经形态计算、可塑性规则验证。触发词：神经形态计算、突触可塑性、钙动力学、STC假说、BrainScaleS、脉冲神经网络、analog neuromorphic、synaptic plasticity、calcium dynamics、spiking neural network。
原文の言語: 中国語
Methodology from Anthropic Economic Index report "Cadences" (Jun 26, 2026) for analyzing AI usage patterns through privacy-preserving telemetry — temporal cadences, output artifact classification, and perception surveys linked to behavioral data.
原文の言語: 英語
Methodology from Anthropic research (May 7, 2026) outlining focus areas for The Anthropic Institute — AI safety research priorities, policy engagement, and societal impact initiatives.
原文の言語: 英語
Mixed-Criticality (MC) system WCET (Worst-Case Execution Time) analysis and optimization using AnTi-MiCS and MulTi-MiCS frameworks. Enables optimal low WCET determination for real-time embedded systems, balancing processor utilization against…
原文の言語: 英語
Generalized framework of antisymmetric cross-polyspectral indices for identifying high-order neural interactions. Quantifies cross-frequency coupling while being intrinsically robust to volume conduction artifacts. Applicable to EEG/MEG analysis and…
原文の言語: 英語
Scalable framework for application-level quantum benchmarking. Supports cross-platform comparison with 13 benchmark families covering solution quality, execution time, energy, and Time-to-Solution metrics. Activation: quantum benchmarking, application-level…
原文の言語: 英語
Approximate quantum error correction theory for non-isometric codes addressing finite-energy experimental realizations
原文の言語: 英語
Self-supervised spiking neural network framework for estimating per-pixel optical flow from event camera streams in underwater environments. Bridges neuromorphic sensing and aquatic intelligence for lightweight, real-time, low-cost perception on…
原文の言語: 英語
Multi-scale brain simulation framework integrating microscopic neural dynamics (Arbor) with macroscopic whole-brain models (The Virtual Brain/TVB). Enables bidirectional real-time co-simulation between detailed spiking neurons and large-scale brain networks.…
原文の言語: 英語
ARFIMA decomposition of stride-to-stride fluctuations in human walking for sensorimotor control analysis. Activation triggers: stride fluctuations, human gait, DFA, ARFIMA, fractal analysis, sensorimotor control
原文の言語: 英語
联合突触前短时程可塑性方法论。扩展Fisher信息学习规则到Tsodyks-Markram突触，推导基线权重和释放概率的学习规则。触发词：突触可塑性、短时程可塑性、突触前、信息论、Tsodyks-Markram、学习规则、temporal coding、STP、synaptic plasticity、presynaptic。
原文の言語: 中国語
ASTDP-GAD: Neuromorphic Graph Anomaly Detection via Adaptive Spiking Temporal Dynamics Plasticity and Spiking Graph Neural Networks. Integrates STDP learning with spiking GNNs for energy-efficient anomaly detection in dynamic networks. Use when building…
原文の言語: 英語
ASTDP-GAD: Neuromorphic Graph Anomaly Detection via Adaptive STDP and Spiking Graph Neural Networks. Integrates adaptive spiking temporal dynamics plasticity with graph anomaly detection for energy-efficient neuromorphic deployment.
原文の言語: 英語
Astrocyte-centric 3-body plasticity framework — how astrocytes participate in synaptic credit assignment alongside pre- and postsynaptic neurons. Proposes astrocytes as slow modulatory agents that bridge Hebbian millisecond-timescale plasticity with memory…
原文の言語: 英語
Asymmetric nonlinear return extrapolation framework for optimal portfolio choice under stochastic volatility. Extends return extrapolation with saturation in belief updating and gain/loss asymmetry. Use when: behavioral portfolio optimization, return…
原文の言語: 英語
Asynchronous Delta Modulation (ADM) for Brain-Machine Interface (BMI) applications. Converts continuous neural signals into event-driven spike trains using adaptive delta modulation, enabling ultra-low-power, low-latency neural encoding for implantable and…
原文の言語: 英語
Asynchronous quantum distributed computing - implementing global quantum operations in distributed systems. Combines classical distributed algorithms (Chandy-Lamport) with quantum computing principles. Activation: quantum distributed, quantum snapshot,…
原文の言語: 英語
Atlas-free Brain Network Transformer using individualized functional brain network (iFBN) framework. Constructs subject-specific brain networks via spatial ICA components as nodes and functional connectivity as edges. Eliminates atlas dependency. Activation:…
原文の言語: 英語
Universal EEG representation learning using microstate tokenizers. Builds a discrete microstate tokenizer from large-scale EEG datasets by clustering continuous signals into sequences of quasi-stable brain activity patterns. Outperforms traditional…
原文の言語: 英語
ATP滞后现象三方突触建模方法论。星形胶质细胞释放ATP累积为腺苷，通过A1受体产生滞后反馈抑制。适用于神经胶质相互作用、突触可塑性、三方突触建模。触发词：ATP滞后、三方突触、星形胶质细胞、腺苷、突触可塑性、tripartite synapse、astrocyte、adenosine。
原文の言語: 中国語
Attention to task structure for cognitive flexibility — neural mechanisms enabling flexible switching between task rules. Demonstrates how attentional mechanisms gate task-relevant information for cognitive control. Applicable to cognitive neuroscience,…
原文の言語: 英語
神经网络吸引子元动力学方法论。研究慢适应过程如何塑造吸引子景观演化。适用于神经动力学、连续学习。触发词：吸引子、元动力学、神经动力学、attractor、metadynamics。
原文の言語: 中国語
Black-box evaluation framework for assessing LLM-generated Design Structure Matrices (DSMs) from structured technical documentation. Integrates structural metrics (Completeness, Correctness, Coupling Density), classification metrics (Selective Accuracy,…
原文の言語: 英語
Paper analysis: AutoCog — an automated LLM-driven system for discovering cognitive theories. The system takes existing theories as seeds, generates novel hypotheses, tests them against data, and produces executable cognitive models. Validated with human…
原文の言語: 英語
Automated neural characterization using natural language and digital twins. Closed-loop framework that translates neuron activation patterns into concise semantic descriptions, generates hypothesis images, and verifies them in silico. Use when studying:…
原文の言語: 英語
Autopoiesis paradigm for self-evolving systems - online policy evolution for LLM serving under runtime dynamics. LLM-driven program synthesis for continuous adaptation. Activation: self-evolving systems, adaptive serving, online policy evolution, LLM serving…
原文の言語: 英語
Framework for probabilistic prediction of neural population dynamics using autoregressive flow matching (AFM). Addresses the inherent stochasticity and nonlinearity of neural activity by leveraging transport-based generative modeling to forecast neural…
原文の言語: 英語
Autonomous AI research loop - let the agent run ML experiments overnight. Inspired by Karpathy's autoresearch. Use when: autonomous research, ml experiments, overnight training, self-improving models, auto-optimization.
原文の言語: 英語
Methodology for analyzing the misalignment between backpropagation algorithms in deep neural networks and the hierarchical organization of brain responses. Extends encoding analyses from forward activations to backpropagated gradients, revealing fundamental…
原文の言語: 英語
反向传播算法与人脑视觉处理层级的不匹配研究。使用fMRI和MEG证明梯度虽能预测脑信号，但其时空组织与生物学反向传播机制不符。激活词：反向传播、大脑层级、brain hierarchy、backpropagation、fMRI、MEG、视觉处理、DINOv3、神经网络学习机制。
原文の言語: 複数言語
Backpropagation algorithm misalignment with human brain visual processing hierarchy research. Uses fMRI/MEG to map backpropagated gradients onto neural data, showing DINOv3 gradients can predict brain signals but spatial/temporal organization diverges from…
原文の言語: 英語
Adaptive frequency-aware neural network for EEG artifact removal using band-specific processing and routing mechanisms. Activation triggers: EEG denoising, artifact removal, BandRouteNet, EOG removal, EMG removal.
原文の言語: 英語
BandRouteNet adaptive frequency-aware neural network for EEG artifact removal. Jointly exploits band-specific processing and full-band contextual modeling for EOG/EMG artifact denoising. Features band-wise denoising with routing mechanism and full-band…
原文の言語: 英語
Bandwidth reduction methods for packetized Model Predictive Control over lossy networks. Multi-horizon MPC formulation with communication-rate reduction for networked control systems. Use for: networked MPC, bandwidth-efficient control, 5G/IoT control…
原文の言語: 英語
BaRA — Bayesian Adaptive Rank Allocation for LoRA fine-tuning. Dynamically allocates per-instance effective rank via sparse activation of disentangled latent factors, with complexity-theoretic generalization bounds depending on learned joint effective rank…
原文の言語: 英語
Barbell Codes methodology for implementing qLDPC error correction on superconducting quantum hardware with constant hardware complexity scaling.
原文の言語: 英語
Bayesian adaptive latent mixture model for zero-inflated weighted brain connectome analysis. Use when analyzing structural/functional brain networks with many zero-valued edges, modeling subject-level mixture of shared connectivity templates, or performing…
原文の言語: 英語
Bayesian dynamical framework for modeling time-order effects in sequential haptic perception. Captures perceptual biases from prior expectations and temporal structure using drift-diffusion dynamics. Activation: haptic perception, Bayesian dynamics,…
原文の言語: 英語
Bayesian framework for quantifying neural entrainment evidence in Information Processing Pathway Maps (IPPMs). Shifts from frequentist hypothesis testing to probabilistic model adjudication, enabling relative evidence quantification for competing…
原文の言語: 英語