Menu
Skills in this repository
SkillsMP has collected 4,114 skills from hiyenwong/ai_collection. Open a skill to review its source and details.
hiyenwong/ai_collectionShowing 40 of 4,114 collected skills.
Just EEG Transformer (JET) — generative EEG framework using conditional flow matching to model neural signals as continuous trajectories, preserving spectral structure, temporal stationarity, and signal statistics. ICML 2026. Reduces TS-FID by >40% on…
Non-monetary karma economy for fair distributed resource allocation. Online karma auctions for EV charging, distributed scheduling, and capacity management. Use when: fair resource allocation, non-monetary economies, distributed scheduling, EV charging…
KAST-BAR methodology: Knowledge-Anchored Semantically-Dynamic Topology Brain Autoregressive Modeling for universal neural interpretation. Integrates Dual-Stream Hierarchical Attention (DSHA) encoder for brain topology, Knowledge-Anchored Semantic Profiler…
End-to-end academic research workflow using knowledge graphs. Searches papers from arxiv/web, imports to KG database, generates embeddings, runs graph algorithms (PageRank, vector search), and extracts patterns for skill creation. Use for: automated research…
Zero-shot handwriting BCI decoding via conserved kinematic representations. Aligns neural activity to imagined kinematics for open-vocabulary character decoding without per-character training data. Activation: zero-shot BCI, handwriting decoding, kinematic…
Efficient event-driven retrieval in high-capacity kernel Hopfield networks. Asynchronous sequential updates achieve statistically indistinguishable trajectories from synchronous dynamics with storage capacity P/N ≈ 30. Suitable for low-power neuromorphic…
Koopman-von Neumann (KvN) molecular dynamics methodology for computing Green-Kubo transport coefficients as quantum algorithm readout problems. Formulates classical NVE and NVT dynamics as unitary evolutions on Hilbert spaces, enabling quantum speedup for…
Theoretical framework demonstrating that mean-field chaos in random recurrent networks is predictable from continuous past history
Source text: Chinese
Kuramoto模型脑网络相位动力学分析方法论。使用振荡器同步框架研究脑网络相位耦合，分析催产素等神经调节物质对脑网络动态的影响。适用于脑网络同步性分析、神经调节研究、网络神经科学。触发词：Kuramoto模型、脑网络、相位耦合、同步性、神经调节、催产素、oxytocin、phase coupling、synchronization、brain network dynamics。
Source text: Chinese
Unified control-theoretic framework for complex-valued Kuramoto networks. Based on arxiv:2604.07249 'Complex-Valued Kuramoto Networks: A Unified Control-Theoretic Framework' by Giordano et al. Use when analyzing Kuramoto network synchronization, phase locking…
Kuramoto-von Mises时间序列模型用于耦合振荡器的概率建模。无需假设热力学平衡，通过Langevin动力学构造实现非平衡 regime的准确建模，在高采样率下具有闭式代数解。
Source text: Chinese
L-SPINE low-precision SIMD spiking neural compute engine with unified multi-precision datapath (INT2/4/8). Multiplier-less shift-add model for FPGA-based edge SNN inference with 3 orders of magnitude energy efficiency improvement. Activation: L-SPINE, SIMD…
LACUNA testbed methodology for evaluating LLM unlearning localization precision. Use when assessing whether unlearning truly erases knowledge from model parameters or merely obfuscates it, benchmarking unlearning methods, detecting resurfacing attacks, or…
Large fluctuation theory for open quantum systems — analyzing atypical measurement outcomes in driven dissipative steady states. Shows large-deviation functions develop lines and surfaces with discontinuous derivatives, unlike equilibrium analytic Wigner…
Comprehensive survey of machine learning methods for studying latent neural activity dynamics - from state-space models to deep generative models covering single-region dynamics, multi-region communication, and neural manifold geometry.
LatentRevise — First-order latent revision method that recovers training signal from zero-hit prompts in RLVR. Optimizes input embeddings of failed reasoning prefixes under dual gradients (away from failed continuation, toward gold answer), constrained to…
Lateral predictive coding (LPC) framework for feature detection in biological neural circuits with modular network structures. Analyzes response time trade-offs between modular vs non-modular architectures. Activation: predictive coding, neural circuits,…
Lattice Field Theory (LFT) framework for interpreting BCI recordings from real neural networks. Applies physics-based formalism to neural data analysis, connecting Maximum Entropy models with Free Energy Principle.
Lattice surgery methodology for fault-tolerant logical operations between distance-three surface-code logical qubits on planar superconducting processors. Enables deterministic Bell state preparation, two-qubit Deutsch-Jozsa at logical level, and magic-state…
Laya: A LeJEPA approach to EEG via Latent Prediction over Reconstructed Activity. Self-supervised EEG foundation model using Joint Embedding Predictive Architecture (JEPA) with latent-space prediction for EEG representation learning. Enables transfer learning…
4D and 5D Layer Codes through Color Routing — CSS code construction generalizing Layer codes to d dimensions using qLDPC embedding and color routing. Saturates d-dimensional BPT bounds exactly, modular architecture for network patches. Activation: layer…
Learnable Observable Quantum Neural Network methodology. Makes quantum measurement observables (Hermitian matrices) trainable parameters alongside circuit parameters. Reduces expert dependency in QML. Activation: quantum observable, QNN measurement, learnable…
LearnAD: neuro-symbolic method for Alzheimer's disease classification from brain MRI, learning fully interpretable rules combining statistical models and decision logic. Activation: LearnAD, Alzheimer interpretable rules, neuro-symbolic AD classification.
Learning rich neuron dynamics within deep Spiking Neural Networks (SNNs). Moves beyond simple LIF neuron models to capture complex temporal dynamics in deep SNN architectures. Enables more biologically realistic and computationally powerful spiking networks.…
Comparative methodology for analyzing brain alignment across learning rules (backpropagation, feedback alignment, predictive coding, STDP). Tracks representational similarity analysis (RSA) alignment to human fMRI data during training. Key finding: local…
Learning sequence timing and control of replay speed in networks of spiking neurons. Extends the spiking Temporal Memory (sTM) model to encode element-specific duration and flexibly control replay speed via oscillatory background inputs. Applicable to…
Learning sequence timing and control of replay speed in networks of spiking neurons — sTM model extension for encoding element-specific timing and flexible replay speed modulation via oscillatory background input. arXiv 2605.22523 (May 2026).
Leggett-Garg inequality testing methodology for neural dynamics. Proposes experimental tests to distinguish diffusive vs non-diffusive stochastic structure in single neurons using temporal correlations. Connects Kac processes to Telegrapher equation and…
Exact variance and Fano factor analytical formulae for arbitrary level crossings in stationary Gaussian processes. Extends the Kac-Rice mean crossing rate to capture clustering vs. regularity statistics, critical for neuronal spike train analysis, neural…
开源 LFSR 随机泄漏积分-发放神经元硬件实现方法论 — SkyWater 130nm CMOS 工艺上的随机脉冲神经网络神经元设计
Source text: Chinese
Layer-wise Interactive Dual-Stream Network (LI-DSN) for EEG Motor Imagery decoding with cross-subject generalization, spatial-temporal dual-stream architecture, and layer-wise interactive fusion. Addresses the challenge of EEG signal variability across…
Hardware-aware quantum circuit synthesis using Lie group diffusion models on SU(2) manifold. Combines discrete skeleton selection with continuous gate parameter generation via heat kernel denoising. Use when: compiling quantum circuits, synthesizing…
Sample complexity analysis for stabilizer state testing and learning under k-qubit coherent memory constraints. Establishes fundamental memory-information tradeoffs in quantum information processing.
Sample complexity analysis methodology for quantum Lindbladian simulation using Wave Matrix Lindbladization (WML) algorithm. Provides explicit non-asymptotic bounds, dimension dependence analysis, and typical-case guarantees for random Lindblad operators.…
脑结构-功能耦合线性生成框架。从扩散加权成像(DWI)结构连接预测静息态fMRI功能连接，揭示整合枢纽和中介枢纽的机制，支持虚拟损伤实验。适用于脑连接组学、神经影像分析、结构-功能关系研究。触发词：结构功能耦合、脑连接组、结构连接、功能连接、整合枢纽、structure-function coupling、structural connectivity、functional connectivity、integrator hub、mediator hub。
Source text: Chinese
LiquidTAD: Efficient temporal action detection via liquid neural dynamics. Replaces self-attention with parallelized ActionLiquid blocks for parameter-efficient video understanding. Use when building temporal action detection systems, deploying on edge…
LLM Agent工具盲从现象研究。当LLM agent配备GNN工具时，agent不判断工具输出，而是盲目服从。更强的LLM backbone反而defer更多。
Source text: Chinese
Agentic Large Language Model framework for active Fault-Tolerant Control (FTC) in cyber-physical systems. Combines multi-agent LLM workflow with Digital Process Plant Twin and Graph RAG (CPSMod ontology) to transform fault detection outputs into…
Autonomous LLM agent methodology for computational physics discovery using progressive local search, knowledge accumulation from successful/failed attempts, and interpretable exploration trajectories. Covers PhyNex framework for scorable scientific tasks with…
LLM中的心理概念神经元识别与控制方法论。使用探针分析识别大五人格(Big Five)概念神经元，通过干预增强/抑制其激活来控制LLM生成偏向。揭示表示控制与行为控制之间的差距。Activation: LLM concept neurons, neural control, Big Five, representation steering, activation intervention, psychological constructs.
Source text: Chinese