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.
Non-equilibrium thermodynamic framework linking macroscopic predictive performance of quantum reservoir computing to microscopic energetic costs. Maps Holevo capacities onto Bogoliubov-Kubo-Mori geometric manifold. Identifies spectral resonance at quantum…
Methodology for designing and evaluating AI scientific capabilities through domain-specific benchmarks. Covers BioMysteryBench design principles, multi-step reasoning evaluation, and human-expert comparison methodologies.
Hamiltonian-informed diagnostic benchmark for Quantum Architecture Search (QAS). Organizes molecules into structural tiers via Pauli operator fingerprints, computational basis representation, and ground-state entanglement. Detects failure modes invisible to…
L-PACT (Locked Predictive-Aligned Cross-modal Testing) framework for rigorous brain-language model alignment evaluation. Goes beyond prediction scores with four evidence gates: predictive-control, relational-profile, mechanism-stripping, and…
Natural language hypothesis generation and verification for single-neuron selectivity. Combines vision-language models, neural digital twins, and text-to-image generation to automatically characterize what individual neurons encode across the visual…
NeuralBench unified benchmarking framework for NeuroAI models. Standardized evaluation across EEG/MEG/fMRI tasks with 36 tasks, 14 architectures, 94 datasets. Covers foundation model evaluation, task-specific baselines, cross-modal extension. Activation:…
Apply sparse autoencoders to analyze internal representations of neural quantum states and steer quantum properties.
Benchmark design methodology for evaluating AI scientific capabilities in open-ended, generative research contexts. Covers qualitative data collection, longitudinal tracking, and expectation measurement.
Probabilistic memory (p-MEM) — unified memory primitive for trustworthy edge intelligence that stores distribution parameters and samples at native memory bandwidth
Retrieval-Based Brain Decoding by Alignment, not Complexity. Linear contrastive decoders outperform ridge regression and non-linear alternatives across images, text, and sound. Decoding gains arise from training objective choice, not architectural complexity.
Minimum-Distortion Embedding (MDE) framework for analyzing evolving neuronal network dynamics. Use when dimensionality-reducing high-dimensional spiking activity, analyzing network development trajectories, or comparing stimulation effects in neuronal…
CNN + Adversarial Autoencoder (AAE) for EEG signal classification — from raw EEG to image representations, latent-space regularization, and robust brain-computer interface (BCI) decoding.
Canonical quantization methodology for constructing quantum neuron models from classical Hamiltonians — a principled framework for quantum machine learning primitives
Constraint-preserving QAOA using XY-mixer for multiplex CRISPR gene editing optimization. Systematically compares structural constraint enforcement (XY-mixer) vs penalty-based approaches across simulator and real hardware. XY-mixer achieves >95% optimum…
Graph-based Policy Optimization (GraphPO) for reasoning models. Represents rollouts as DAGs, merges semantically equivalent paths, and improves advantage estimation variance through graph structure.
Vision SmolMamba: Spike-Guided Token Pruning for energy-efficient spiking state-space vision models. Combines SNN event-driven sparsity with Mamba selective recurrence via SST-TP (Spike-Guided Spatio-Temporal Token Pruner). Activation: Vision SmolMamba,…
Unified multi-modal framework integrating PPO robo-advisory, HFT prediction, in-context investment advisory, game-theoretic banking, and cross-modal sentiment analysis. Use when: unified financial AI systems, multi-domain financial AI, robo-advisory…
Disagreement-Modulated Policy Self-Distillation framework for LLM reasoning. Resolves privileged information leakage and exploration preservation in on-policy distillation via reverse-KL barycenter target, achieving leakage attenuation and exploration…
Weak-to-strong generalization methodology transferring RL-induced policy shifts as dense implicit reward signals from smaller to larger models, enabling cross-scale RL outcome reuse.
Conserved Kinematic Representations for Zero-Shot Decoding in Handwriting BCIs. Methodology aligning neural activity to imagined kinematics for zero-shot capable ML decoding of unseen characters in BCI systems. Use when: researching brain-computer interfaces,…
Kirchhoff-Inspired Neural Network (KINN) - state-variable-based network architecture built on Kirchhoff's current law for evolving high-order perception. Derives numerically stable state updates from ODEs, enabling explicit decoupling and encoding of…
Meta-learning methodology for achieving human-like visual representations. Proposes that meta-learning (learning to learn) pressure shapes neural representations to support open-ended tasks. Compared to pretrained models, meta-learned representations better…
元学习上下文方法实现无需训练的跨被试脑解码。通过上下文学习实现训练无关的跨个体fMRI解码。适用于零样本脑解码、快速脑机接口、个体化神经科学。触发词：元学习脑解码、上下文学习、跨被试、训练无关、零样本。
Source text: Chinese
Meta-learning In-Context approach for training-free cross-subject brain decoding. Enables zero-calibration BCI through context-based meta-learning. Triggers: meta-learning, brain decoding, cross-subject, training-free, in-context learning, zero-calibration…
Meta-learning in-context brain decoding methodology for zero-shot cross-subject generalization in BCI. Enables training-free adaptation to new users by framing brain signal decoding as an in-context learning problem — constructing support sets from other…
BrainCoDec v4 — Foundation framework for training-free cross-subject fMRI-based semantic visual decoding via meta-optimized in-context learning. Achieves zero-shot generalization across subjects and scanners without anatomical alignment or stimulus overlap.…
BrainCoDec v5 — Foundation framework for training-free cross-subject brain decoding using meta-learning in-context approach. Enables zero-shot visual decoding from fMRI without subject-specific training. Activation: brain decoding, meta-learning, in-context,…
Multi-Plasticity Continual System (MPCS) integrating 11 neuroplastic mechanisms for continual learning. Key finding: EWC regularization degrades performance at high task similarity. Pareto frontier analysis for model compression. Activates: continual…
Nonlinear separation principle for recurrent neural networks (RNNs) using contraction theory. Guarantees global exponential stability for contracting state-feedback controllers and observers. Applies to firing-rate and Hopfield RNN architectures. Based on…
OPD-Evolver - On-policy distillation framework for cultivating holistic agent evolvers. Slow-fast co-evolution with four-level memory hierarchy: read, use, write, maintain experience. Outcome-calibrated memory attribution + privileged hindsight distillation.…
Optical neural networks using coherent transient dynamics in waveguide QED for all-optical neuromorphic computing
Partial fusion of neural networks interpolating between ensembles and weight aggregation via neuron-level similarity matching and partial optimal transport. Frames partial fusion as generalized pruning where neurons are deleted or linearly combined.
Phenomenological Renormalization Group (PRG) validation methodology for detecting criticality in neuronal models. Validates PRG coarse-graining on excitable cellular automata and stochastic E/I LIF networks, introduces adaptive ISI-based time binning to…
phys-MCP: substrate-aware control plane architecture for heterogeneous Physical Neural Networks (PNNs) spanning molecular, chemical, biological, photonic, memristive, and mechanical substrates. Provides capability models, lifecycle semantics, telemetry…
Design neural networks that embed physical constraints (equations, symmetries, conservation laws) directly into the computational graph. Use when modeling physical systems, scientific computing, or when physics-informed AI is needed. Keywords: PGNN,…
Physics-guided neural network design and training methods. Embed physical laws, constraints, and symmetries into neural network architecture for improved modeling of physical systems (quantum mechanics, statistical physics, fluid dynamics, materials science).…
Quantum qutrit-based neural network methodology for real-time financial forecasting. Uses 3-state quantum neurons instead of 2-state qubits to capture richer financial patterns with faster training.
Renormalization group (RG) framework for analyzing scaling laws and criticality in brain activity. Connects 1/f noise, neuronal avalanches, and coarse-grained descriptions through RG theory. Activates: renormalization brain, scaling law neural activity, 1/f…
ReOPD (Replayed-Prefix On-Policy Distillation) methodology for scalable multi-turn agent distillation without environment interaction during training. Addresses the 'prefix trap' in multi-turn OPD via reliability-aware prefix sampling.
SABER framework integrating spatial attention neuroscience with Extended Reality for adaptive human-computer interaction. Activation: spatial attention XR, brain-computer interface, attention-aware computing, extended reality neuroscience, eye-tracking…