hiyenwong/ai_collection - Page 16

claude-code

Delegate coding to Claude Code CLI (features, PRs).

Langue du texte source : anglais

claw-r1-agentic-rl-data-middleware

Step-level data middleware system for agentic RL. Captures multi-turn agent-environment interactions and organizes them as managed data assets.

Langue du texte source : anglais

clockless-asynchronous-neuromorphic-computing

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…

Langue du texte source : anglais

clockless-neuromorphic-snn

Clockless (asynchronous) digital Boolean spiking neural networks for neuromorphic computing. Based on arXiv:2605.16114 (May 2026). Use when: designing clockless/async neuromorphic hardware, implementing Boolean spiking neurons on FPGA, liquid state machines…

Langue du texte source : anglais

closed-form-predictive-coding-hgf

Closed-form predictive coding via hierarchical Gaussian filters (HGF) methodology from arXiv:2605.20293. Restores precision-weighted prediction errors to predictive coding networks by expressing them as deep hierarchical Gaussian filters. Enables biologically…

Langue du texte source : anglais

closed-loop-quantum-probabilities-unitarity

Closed-loop decomposition of quantum probabilities as direct consequence of unitarity, identifying Bargmann invariants as phase-invariant loop quantities and Born rule as quadratic forward-reverse amplitude product. arXiv:2606.02504

Langue du texte source : anglais

clp-snn-loihi2-continual-learning

Online Continual Learning on Intel Loihi 2 via a Co-designed Spiking Neural Network (CLP-SNN). Covers self-normalizing local learning rules, spike-driven neural state machines for autonomous on-chip learning, and breakthrough efficiency gains on neuromorphic…

Langue du texte source : anglais

clsa-cross-layer-sparse-attention

Cross-layer sparse attention sharing routing index across decoder layers for 7.6x decoding speedup and 17.1x throughput improvement at 128K context

Langue du texte source : anglais

cmos-invertible-logic-stochastic-computing

CMOS invertible logic using spiking stochastic computing methodology. Implements bidirectional (forward/reverse) logic gates via Boltzmann machines with simple spiking neural networks. Activation: invertible logic, stochastic computing, CMOS neuromorphic,…

Langue du texte source : anglais

cmos-nonlinear-classification-biologically-realistic

Biologically Realistic Dynamics for Nonlinear Classification in CMOS+X Neurons. CMOS+X technology for realizing biologically realistic nonlinear neuronal dynamics for efficient spiking neural network classification. Activation: CMOS+X neurons, nonlinear…

Langue du texte source : anglais

cmut-transcranial-ultrasound-bbb

Capacitive Micromachined Ultrasonic Transducer (CMUT)-based transcranial focused ultrasound system for blood-brain barrier opening and drug delivery. Includes phase-inversion transmission, microbubble monitoring, and closed-loop control. Activation: CMUT…

Langue du texte source : anglais

cnn-snn-eeg-imagined-speech

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,…

Langue du texte source : anglais

cnn-snn-imagined-speech-decoding

EEG-based imagined speech decoding using hybrid CNN-SNN architecture. First integration of spiking neural networks for imagined speech BCI, achieving 80.13% accuracy on BCI Competition III benchmark. Covers CNN feature extraction, SNN temporal classification,…

Langue du texte source : anglais

coarse-feedback-human-aligned-visual

Shows that extremely coarse feedback signals (distinguishing as few as 8 broad categories) produce neural representations matching or exceeding 1000-class supervised models in brain alignment with primate vision. Trains hundreds of CNNs and ViTs across…

Langue du texte source : anglais

code-modulated-motion-vep-bci

Code-Modulated Motion Visual Evoked Potential (c-MVEP) methodology for brain-computer interfacing using motion stimulation instead of flickering. Use when: designing BCI paradigms, visual evoked potential stimulation, motion-based BCI, reducing visual fatigue…

Langue du texte source : anglais

codex-session-manager

Monitor and manage OpenAI Codex CLI sessions. List history, view session content, resume via ACP, track statistics. Use when user asks about Codex sessions, wants to check progress, or resume previous work.

Langue du texte source : anglais

coding-agents-social-sciences-research

Methodology from Anthropic economic research (May 27, 2026) on using AI coding agents to accelerate social science research workflows — experimental design, data analysis, literature review, and reproducibility.

Langue du texte source : anglais

coflow-scheduling-ocs

Coflow Scheduling in Multi-Core Optical Circuit Switching Networks with Performance Guarantees

Langue du texte source : anglais

cognisnn-brain-inspired-snn

CogniSNN (Cognition-aware Spiking Neural Network) is a comprehensive framework that bridges artificial intelligence and computational neuroscience through Random Graph Architecture (RGA). It implements three core brain-inspired mechanisms:…

Langue du texte source : anglais

cognisnn-random-graph-snn

CogniSNN: Cognition-aware Spiking Neural Network with Random Graph Architecture enabling neuron-expandability, pathway-reusability, and dynamic-configurability. Uses Key Pathway-based Learning (KP-LwF) for multi-task transfer and Dynamic Growth Learning (DGL)…

Langue du texte source : anglais

cognition-inspired-dual-stream-emotion

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.…

Langue du texte source : anglais

cognitive-flexibility-task-structure

Neural network model of attention to task structure enabling cognitive flexibility. Studies how neural networks learn to attend to relevant task dimensions and switch between task rules. Combines attention mechanisms with structured task representations to…

Langue du texte source : anglais

cognitive-load-multiscale-attractors

Integrating Cognitive Load and Embodied Cognition Theories Through Representations as Multi-Scale Attractors — proposing a formal rapprochement between cognitive load theory and embodied cognition by reconceptualizing psychological representations as dynamic…

Langue du texte source : anglais

coherence-law-noisy-equivariant-qnn-trainability

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…

Langue du texte source : anglais

cold-atom-medical-imaging

Medical imaging classification using cold-atom (neutral-atom) reservoir computing with auto-encoders and surrogate-driven training. Based on arXiv:2605 (2026). Use when building hybrid quantum-classical pipelines for medical image analysis, applying…

Langue du texte source : anglais

combinatorial-complex-brain-fmri

The Human Brain as a Combinatorial Complex - framework for constructing combinatorial complexes from fMRI time series that captures both pairwise and higher-order neural interactions through information-theoretic measures. Bridges topological deep learning…

Langue du texte source : anglais

common-synaptic-input-estimation-hdemg

Practical methodology for estimating common synaptic input to spinal motor neurons from high-density surface EMG motor unit spike trains using openhdemg. Three complementary approaches: time-domain, frequency-domain, and network-information methods.

Langue du texte source : anglais

commuting-pauli-parallelization

Methodology for optimizing parallel execution of commuting Pauli Product Rotations in fault-tolerant quantum computation. Use when: (1) compiling quantum programs to Pauli Product Measurements (PPMs), (2) reducing circuit depth in surface code architectures…

Langue du texte source : anglais

compactionrl-context-compaction-agents

CompactionRL methodology for training long-horizon agentic LLMs with context compaction. Jointly optimizes task execution and summary generation with token-level loss normalization and cross-trajectory GAE.

Langue du texte source : anglais

competition-stability-ei-circuits

Game-theoretic energetic framework for excitatory-inhibitory neural circuits - competition, stability, and functionality in asymmetric networks

Langue du texte source : anglais

compiler-world-model-tensor-optimization

World-model-inspired evaluator for tensor program optimization. Models schedule evaluation as action-conditioned latent dynamics over program states.

Langue du texte source : anglais

complex-berry-phase-quantum-control

Complex Berry phase measurement and control methodology for non-Hermitian quantum systems. Experimental measurement of real and imaginary Berry phase components using superconducting transmon circuits with engineered dissipation. Path-dependent effects enable…

Langue du texte source : anglais

complex-brain-hypothesis

Complex Brain Hypothesis (CBH) methodology for resolving entropy-content conundrum in consciousness research. Extends Entropic Brain Hypothesis by introducing brain complexity as index of phenomenal richness, modulated by inference grain. Use when: studying…

Langue du texte source : anglais

complex-system-robustness-collapse

Complex system robustness and collapse analysis - temporal structure, percolation methods, phase transitions, bistability, catastrophic collapse. Activation: system robustness, system collapse, complex network, phase transition, resilience analysis.

Langue du texte source : anglais

complex-valued-gnn-control

Complex-Valued GNN Control - 复数值图神经网络用于分布式基不变控制系统。核心技术：复数域几何表示、相位等变激活、全局基不变性。适用于GPS拒绝环境分布式控制。激活词：complex GNN, 复数值 GNN, basis invariant, 基不变控制.

Langue du texte source : chinois

complex-valued-kuramoto-control

Complex-valued Kuramoto network synchronization control using switched control and sliding-mode methods. Embeds phase dynamics into linear state space for tractable control design. Use for: coupled oscillator networks, phase synchronization, Kuramoto model…

Langue du texte source : anglais

complex-valued-neuromorphic-magnitude-phase

Complex-valued neural network with magnitude-phase decomposition for event-driven neuromorphic learning, enabling efficient spiking computation with rich representational capacity

Langue du texte source : anglais

composite-quantum-gates-error-cancellation

Composite quantum pulse gates that simultaneously compensate multiple systematic errors via derivative-based error cancellation. Covers derivative removal by adiabatic gate (DRAG)-extended pulse shaping, multi-error compensating composite sequences,…

Langue du texte source : anglais

compress-distill-reasoning-trace-compression

Post-hoc compression of reasoning chain-of-thought traces before knowledge distillation for efficient training and inference

Langue du texte source : anglais

compressed-minimum-purity-time-evolution

CoMPuTE method for late-time quantum dynamics simulation using compressed minimum-purity time evolution. Tracks reduced local density matrices via minimum-purity principle for efficient long-time quantum many-body simulations.

Langue du texte source : anglais