All categories
Code Review & Quality
Code review, refactoring, debugging, maintainability, and engineering quality
- 16,459
- 686
Security grades appear on each card once the skill has been scanned. Newly imported skills may briefly show without a grade until the backfill job runs.
Open in full browserBrowse code review & quality skills
Showing 8,857–8,880 of 16,459 skills
- Working Memory Rsnn DelaysWorking memory implementation in recurrent spiking neural networks using heterogeneous synaptic delays. Each synapse equipped with D=41 delays modeled as weight tensor W ∈ R^(N×N×D), trained end-to-end with surrogate-gradient BPTT. Network stores arbitrary target spike patterns as sequential chains of overlapping Spiking Motifs. Achieves F1=1.0 on synthetic benchmark (M=16 patterns, N=512 neurons, T=1000 steps). Activation: working memory snn, recurrent spiking memory, heterogeneous delays, s...Votes: 0GitHub stars: 3
- Topology Neural Collapse MonitorMonitoring Neural Training with Topology — Footprint-Predictable Collapse Index using Modular Morse Homology Maintenance (MMHM). Detects representational collapse in neural network embeddings before performance metrics degrade. Activation: neural collapse, representational collapse, topological monitoring, training diagnostics, MMHM, embedding degradation, Morse homology.Votes: 0GitHub stars: 3
- The Unreasonable Effectiveness Of Deep Learning In Artificial Intelligence**arXiv ID:** 2002.04806 **Authors:** Terrence J. Sejnowski **Published:** 2020-02-12T05:25:15Z **Abstract:** Deep learning networks have been trained to recognize speech, caption photographs and translate text between languages at high levels of performance. Although applications of deep learning networks to real world problems have become ubiquitous, our understanding of why they are so effective is lacking. These empirical results should not be possible according to sample complexity in st...Votes: 0GitHub stars: 3
- Tea Nets Cognitive NetworkTEA Nets (Target-Event-Agent Networks) — computational framework combining AI and cognitive network science to extract subjects (Agents), verbs (Events), and objects (Targets) from text. Enables interpretable emotion detection, semantic frame analysis, and linguistic inquiry. Activation: TEA Nets, cognitive network science, semantic network extraction, agent-event-target, text network analysis.Votes: 0GitHub stars: 3
- Taco Tail Aware Credit CalibrationFixes the "Positive-Credit Contamination" failure mode in critic-free RL methods like GRPO.Votes: 0GitHub stars: 3
- Stochastic Quantum Neural NetworkStochastic Quantum Neural Network (SQNN) methodology for modeling neural dynamics using quantum superposition, entanglement, and unitary evolution beyond Von Neumann architecture constraints. Explores neuro-quantum correspondence between artificial neural networks and quantum systems. Use when: quantum neural network design, neuro-quantum modeling, stochastic quantum dynamics in AI, biological brain simulation with quantum computing, neural network architecture beyond Von Neumann. arXiv:2511....Votes: 0GitHub stars: 3
- Spin Wave Superconducting CouplingStrong coupling between propagating spin waves and microwave photons in superconducting resonator-magnetic thin film hybrid circuits. Design methodology for hybrid magnonic quantum systems using YIG-on-substrate integration.Votes: 0GitHub stars: 3
- Scaling Laws And Tradeoffs In Recurrent Networks Of Expressive Neurons**arXiv ID:** 2605.12049 **Authors:** Aaron Spieler, Georg Martius, Anna Levina **Published:** 2026-05-12T12:29:33Z **Abstract:** Cortical neurons are complex, multi-timescale processors wired into recurrent circuits, shaped by long evolutionary pressure under stringent biological constraints. Mainstream machine learning, by contrast, predominantly builds models from extremely simple units, a default inherited from early neural-network theory. We treat this as a normative architectural questi...Votes: 0GitHub stars: 3
- Safety Critical Contextual Control RiemannianSafety-critical contextual control via online Riemannian optimization with world models — Penalized Predictive Control (PPC) framework for provably safe control using black-box simulators and score-based density estimation.Votes: 0GitHub stars: 3
- Riemannian Self Attention Eeg DecodingBures-Wasserstein metric-based Riemannian self-attention network for robust EEG decoding in BCI. Overcomes AIM quadratic dependency and ill-conditioning issues. Activation: riemannian self attention EEG, Bures-Wasserstein EEG decoding, GBWAtt, SPD learning BCI, Riemannian manifold EEGVotes: 0GitHub stars: 3
- Rats Register Attention TransformersRATS methodology for analyzing emergent part-based representations in Register Attention Transformers. Reveals how attention patches develop specialized, reusable structural components through register-based communication. Use when: mechanistic interpretability, transformer internal analysis, register attention, emergent representations, attention patch specialization.Votes: 0GitHub stars: 3
- Quantum Reservoir Forecasting Resource EfficientResource-efficient Quantum Reservoir Computing framework for time-series forecasting. Combines fixed quantum reservoir transformation with post-training quantized classical readout for deployment on edge/limited-memory devices.Votes: 0GitHub stars: 3
- Quantum Neuroscience Analysis量子神经科学跨学科分析方法。将量子计算方法应用于神经科学问题，包括量子神经网络(QNN)用于脑信号分析、量子图神经网络(QGNN)用于脑连接、量子算法优化神经动力学建模。激活关键词: quantum neuroscience, quantum neural network, quantum EEG, quantum brain, 量子神经科学, 量子脑科学, QNN neuroscience.Votes: 0GitHub stars: 3
- Quantum Neural Architecture SearchQuantum Neural Network Architecture Search (QNAS) skill for designing efficient quantum neural networks on NISQ hardware. Uses multi-objective optimization (NSGA-II) to balance accuracy, runtime efficiency, and circuit cutting overhead. Apply when designing quantum neural networks, optimizing hybrid quantum-classical architectures, or searching for Pareto-optimal quantum circuit configurations. Keywords: quantum neural network, QNN, quantum architecture search, variational quantum circuit, an...Votes: 0GitHub stars: 3
- Quantum Graph Neural Drug DiscoveryHybrid quantum-classical drug discovery methodology combining Quantum Graph Neural Networks (QGNN) with Variational Quantum Eigensolver (VQE) for molecular property prediction and lead compound optimization. Use when: designing quantum-enhanced drug discovery pipelines, building hybrid QGNN-VQE architectures for molecular analysis, optimizing lead compounds with quantum eigenvalue solvers, or evaluating quantum advantage in pharmaceutical ML workflows.Votes: 0GitHub stars: 3
- Quantum Circuit Distribution OptimizationJoint qubit leasing and quantum circuit distribution optimization methodology. ILP formulation for multi-QC quantum network resource allocation, NP-completeness analysis, greedy algorithm with local search. Covers qubit allocation, gate execution placement, and inter-QC communication tradeoffs (migration vs teleportation). Activation: quantum circuit distribution, qubit leasing, quantum network optimization, JQLQCD, quantum resource allocation, distributed quantum computing.Votes: 0GitHub stars: 3
- Psm Quantum Memory DistributionProgressive Swapping to the Middle (PSM) protocol for entanglement distribution in quantum networks with imperfect quantum memories. Optimizes entanglement swapping order to minimize decoherence from finite memory coherence times. Presented at 2026 EuCNC and 6G Summit.Votes: 0GitHub stars: 3
- Parallelized Hierarchical Connectome SsmParallelized Hierarchical Connectome (PHC) framework upgrading temporal State-Space Models (SSMs) into spatiotemporal recurrent networks. Maps SSM diagonal core to Neuron Layer and inter-neuronal communication to Synapse Layer, enabling parallel-scan-based brain-scale network simulation with O(T log T) complexity. Use when: brain-scale neural network simulation, spatiotemporal SSMs, connectome-constrained neural modeling, parallel scan recurrent networks, spiking SSMs, or efficient long-seque...Votes: 0GitHub stars: 3
- Neural Sentinel Unified Vision Language Model Vlm For License Plate Recognition With Humanintheloop Continual Learning**arXiv ID:** 2602.07051 **Authors:** Karthik Sivakoti **Published:** 2026-02-04T16:04:15Z **Abstract:** Traditional Automatic License Plate Recognition (ALPR) systems employ multi-stage pipelines consisting of object detection networks followed by separate Optical Character Recognition (OCR) modules, introducing compounding errors, increased latency, and architectural complexity. This research presents Neural Sentinel, a novel unified approach that leverages Vision Language Models (VLMs) to ...Votes: 0GitHub stars: 3
- Nanophotonet Pinn Inverse DesignPhysics-informed AI-driven inverse design framework for nonlinear metasurfaces using hybrid CNN-autoencoder architectureVotes: 0GitHub stars: 3
- Iqp Circuit TrainabilityIQP (Instantaneous Quantum Polynomial-time) circuit methodology for near-term quantum optimization. Use when: designing IQP circuits for Hamiltonian optimization, analyzing connectivity-trainability trade-offs in variational quantum circuits, selecting circuit architectures for NISQ-era optimization, understanding how IQP circuit depth/structure affects optimization performance vs trainability (barren plateaus), implementing penalty-free quantum optimization workflows. Core insight: IQP circu...Votes: 0GitHub stars: 3
- Globally Optimal Snn Parameter ReconstructionGlobally optimal Spiking Neural Network (SNN) training via parameter reconstruction. Extends convexification of parallel feedforward threshold networks to parallel recurrent threshold networks, subsuming parallel SNNs as a structured special case. Eliminates surrogate gradient approximation errors by reconstructing optimal parameters directly. Use when training SNNs, optimizing spiking networks, avoiding surrogate gradient issues, or exploring convex SNN training methods. arXiv: 2605.08022Votes: 0GitHub stars: 3
- Dynamic Neural Manifolds ControlDynamic neural manifolds methodology for flexible closed-loop control on neuromorphic hardware. Uses sensory inputs to modulate heterogeneous inhibition, gain, and transient currents, driving rapid subspace rotations to switch between behaviors and fine-grained trajectory control within them.Votes: 0GitHub stars: 3
- Carve Content Aware RecurrentCARVE (Content-Aware Recurrent with Value Efficiency) methodology — content-aware gating for recurrent models that resolves memory-blind gating, value-axis waste, and enables WY-form chunk-parallel training.Votes: 0GitHub stars: 3