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.
Learned generalization of phase correlation that lifts the fixed Fourier basis restriction to discover unknown transformations between observations. Applicable to image registration, non-rigid deformation, and quantum Hamiltonian eigenstate recovery from…
Neural population decoding methods for analyzing high-dimensional neural recordings. Focuses on decoding cognitive states, working memory, and behavior from population activity using dimensionality reduction and dynamical systems approaches.
Methods for analyzing neural population dynamics including dimensionality reduction, trajectory analysis, and dynamical systems modeling. Covers techniques for understanding how populations of neurons encode information and generate behavior. Use when…
Neural QAOA² methodology: end-to-end differentiable framework for joint graph partitioning and QAOA parameter initialization. Uses generative evaluative network (GEN) with differentiable quantum evaluator for gradient-guided learning. Ranks first on 101/183…
Neural QAOA² methodology - using neural networks for differentiable graph partitioning and parameter initialization in quantum combinatorial optimization. Bridges ML and QAOA for scalable NISQ optimization.
Neural-enhanced optimization framework for quantum architecture embedding problems using Distance Encoder Networks. Solves constrained unit disk problems for neutral atom qubit positioning via modified autoencoder with custom Embedding Loss Function.…
Neural Variational Quantum Linear Solver (NVQLS) - first hybrid quantum-classical operator learning framework using Legendre-Galerkin weak formulation for solving parametric PDEs. Achieves superior accuracy with theoretical computational complexity advantages…
Neural-powered unit disk graph embedding for mapping QUBO problems onto quantum annealer connectivity. Uses graph neural networks to solve the minor embedding problem efficiently.
Neural Receptive Fields via Hyperbolic Geometry
Geometric framework for neural receptive field emergence in scale-free networks. Studies how receptive fields organize and couple with stimulus space embedding without fine-tuning. Activation: receptive field geometry, scale-free networks, stimulus space…
Unified framework synthesizing neural/artificial neural network representation reshaping mechanisms across four paradigms: (1) Embodied VR feedback reshapes motor representations for BCI decoding, (2) fMRI visual question answering decodes reshaped…
神经变异性增强人工神经网络鲁棒性方法论。研究相关性噪声如何改善对抗攻击和自然图像修改的鲁棒性，建立生物学可解释的鲁棒神经网络设计策略。
NeuralSet unified Python framework for Neuro-AI research, harmonizing diverse neural recordings (fMRI, M/EEG, spikes) with deep learning embeddings
Source text: Chinese
Attractor landscape methodology for working memory in neural circuits. Analyzes how persistent neural activity patterns form stable attractor states that encode and maintain information during delay periods. Uses dynamical systems theory, bifurcation…
Research methodology for the intersection of neuroscience and quantum physics/computing. Use when analyzing papers or research at the boundary of brain science and quantum mechanics, including quantum neural networks, quantum memory models of cognition,…
Neural-Symbolic Cognitive Architectures combining neural networks with symbolic reasoning for interpretable, robust AI systems. Enables explicit knowledge representation, logical inference, and learning from both data and rules. Applicable to reasoning…
Neuro-Vesicles framework for dynamical neuromodulation in neural networks. Introduces mobile discrete vesicle population as event-based interaction layer alongside network tensors. Applies to: neuromodulation, dynamic network modulation, spiking networks,…
NeuroAI research roadmap bridging neuroscience and AI - identifies three fundamental capabilities for AI advancement: world modeling, motor control, and biological learning. NSF workshop framework for interdisciplinary research. Keywords: NeuroAI, world…
NeuroAI and Beyond: Bridging Between Advances in Neuroscience and Artificial Intelligence (arXiv:2604.18637). NSF workshop report identifying three fundamental capability gaps in current AI systems from neuroscience perspective. Activation: neuroai,…
Neuro-Anatomically Aware Point Cloud Representation (NeuroAPS-Net) for efficient Alzheimer's disease classification from MRI. Converts T1-weighted MRI into anatomically-informed 2D point clouds with region-aware feature encoding. Activation triggers:…
Neurobiological Craving Signature (NCS) - predicts social craving and social reinstatement from neural activity patterns. Identifies biomarkers for social isolation effects on brain circuits involved in social motivation. Activation: neurobiological craving…
NeuroCogMap framework for mapping cognitive organization in LLMs using neuroscience-inspired methods, linking LLM internal representations to human cortical function
Neurocognitive governance framework for autonomous AI agents based on executive function and inhibitory control principles. Maps human self-governance mechanisms to AI decision-making for safety-critical environments. Activation: neurocognitive governance, AI…
Integrative neurocybernetic modeling framework for large-scale neuroscience. Unifies diverse neural datasets across animals, brain areas, and behaviors through cybernetic principles. Addresses fragmentation in computational neuroscience. Keywords:…
Integrative Neurocybernetic Modeling framework for large-scale neuroscience research. Treats the brain as a controller pursuing latent objectives in closed-loop coupling with body and environment. Bridges fragmented computational neuroscience efforts through…
4D(3D×T)扩散模型用于神经退行性疾病脑解剖结构的纵向生成建模。基于形变的形态测量学。
Source text: Chinese
4D (3D×T) diffusion-based generative framework for modeling neurodegenerative brain anatomy progression. Combines spatial and temporal modeling for longitudinal brain imaging and disease progression prediction. Keywords: neurodegenerative disease, 4D…
4D(3D×T)扩散模型用于神经退行性疾病脑解剖结构的纵向生成建模。基于形变的形态测量学(DBM)和平稳速度场(SVF)。
Source text: Chinese
NeuroFlowNet — cross-modal generative framework using Conditional Normalizing Flow for reconstructing high-fidelity iEEG signals from non-invasive scalp EEG
NeuroGAN-3D methodology for high-fidelity 3D generative super-resolution of resting-state fMRI (rs-fMRI) spatial maps. Uses a GAN architecture to enhance spatial resolution of volumetric functional brain network maps, enabling more precise localization of…
Brain-inspired perceptual compression using evidence-driven neuromimetic principles. Leverages human visual system characteristics for efficient data compression that prioritizes perceptually important information.
神经调节突触可塑性的学习框架。在脉冲神经网络（SNN）中通过梯度下降 训练神经科学启发的可塑性模型，解决在线学习问题。 触发词：突触可塑性、神经调节、脉冲神经网络、在线学习、学习的学习、 synaptic plasticity, neuromodulation, spiking neural network, online learning, meta-learning。
Source text: Mixed languages
Neuromodulation-based control architecture for dynamically reconfiguring rhythmic patterns in central pattern generators (CPGs) with fixed connectivity. Uses targeted neuromodulatory inputs to enable rapid, localized rhythmic transitions without structural…
Neuromorphic Correlates of Artificial Consciousness (NCAC) — theoretical framework merging neuromorphic design with brain simulations for artificial consciousness. arXiv:2405.02370
神经形态持续学习方法用于核电厂工业控制系统(ICS)监测的顺序部署。结合脉冲神经网络(SNN)和在线学习，实现关键基础设施的实时异常检测和安全监控，同时防止灾难性遗忘。适用于关键基础设施保护、工业网络安全、边缘AI。
Source text: Chinese
Neuromorphic energy-aware learning for adaptive deep brain stimulation (DBS). Use when working with spiking neural networks (SNNs) for closed-loop neural control, brain-computer interfaces, or neuromorphic hardware deployment. Covers energy-aware RL reward…
Neuromorphic Energy-Aware Learning for Adaptive Deep Brain Stimulation — co-optimizes stimulation energy and inference efficiency via deep spiking Q-network on neuromorphic hardware. Achieves 45.2% oscillation suppression with 80% charge reduction at 0.52 mW.
Neuromorphic control of flapping-wing micro aerial vehicles using SNNs on resource-constrained ESP32 microcontroller. Hierarchical SNN framework: state estimation + CPG modulation for wing actuation. 36% latency reduction, 18% power reduction vs ANN. First…
End-to-end Spiking Neural Network for 3D LiDAR-based Bird's Eye View object detection in autonomous driving, with neuromorphic energy analysis and learned spike encoding.
End-to-end spacecraft 6-DoF pose estimation using event cameras and BrainChip Akida neuromorphic processor. MobileNet-style keypoint regression on event-frame representations with quantization-aware training (8/4-bit) converted to spiking neural networks.…