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.
AI interviewer methodology for operationalizing Descriptive Experience Sampling (DES) into an explicit, inspectable reasoning architecture. Grounded in the established phenomenological method with co-development by DES originator Russell T. Hurlburt. Use when…
An off switch for dual-use knowledge in AI models
Anthropic Economic Index report: Cadences
A skill summarizing the arXiv paper: A 32-channel event-based bio-signal analog front-end with adaptive delta and pulse frequency encoding arXiv ID: http://arxiv.org/abs/2607.12901v1 Authors: Narayanan Shyam, Saptarshi Ghosh, Giacomo Indiveri Published:…
Skill generated from arXiv paper 2607.14086: Leveraging unlabelled data for generalizable neural population decoding
A skill for conducting EEG/MEG-based brain network analysis based on the arXiv paper "Exploring Brain Networks Using Noninvasive Electrophysiological Measurements: Methods and Applications"
Visual Semantic Decoding of Electrocorticography from Video Stimuli using End-to-End Deep Learning - methodology for decoding visual categories from ECoG using Transformer-based deep learning with mixup augmentation and high-gamma band analysis.
A skill for understanding and applying the methods from the arXiv paper: Toward a mechanistic understanding of inference in visual cortex and diffusion models
Automated workflow for researching recent arXiv papers, generating Hermes skills, syncing to ai_collection repo, updating INDEX.md, creating Obsidian notes, and updating the knowledge graph. Designed for cron jobs with no user interaction.
Best practices for safely accessing arXiv programmatically, avoiding common pitfalls with HTTP, SSL, and automated tools.
Workaround for arxiv-search SSL issues when using python httpx
Skill for EEG/MEG-based brain network analysis covering forward/inverse problems, connectivity measures, and pipelines.
Claude plays robotics
Claude’s values across models and languages
Cluster-based distributed small-signal stability certification for grid-forming inverter networks. Use when: (1) certifying stability of large-scale inverter networks without a fully assembled global model; (2) designing cluster-based or decentralized…
Skill for understanding and applying the CogniSNN framework: a Spiking Neural Network paradigm that incorporates Random Graph Architecture to achieve Neuron-Expandability, Pathway-Reusability, and Dynamic-Configurability for brain-inspired intelligence.
Methodology for modeling the co-evolution of human competence and AI tool reliance as a bistable dynamical system. Analyzes competence collapse thresholds, transparency effects, and agency transfer in human-AI collaboration. Use when designing AI tools,…
A skill for implementing cross-domain recombination inspired by dreaming, based on the paper "Discovery by Dreaming: Cross-Domain Recombination in Artificial Memory" (arXiv:2607.16256). This skill outlines how to implement a LoRA fine-tuning pipeline (DREAMS)…
Methodology for training CNNs with eccentricity-constrained egocentric video data to reveal adaptive information coding that mirrors primate visual system organization, showing differential task-relevance between foveal and peripheral vision.
Eccentricity-constrained CNN training on egocentric video reveals adaptive, task-aligned information coding across the visual field, with fovea-preferred models advantaged for face and object tasks and periphery-preferred models favored in scene-selective…
Eccentricity-Constrained CNN Training methodology for adaptive visual information coding around the visual field using egocentric data
Skill for analyzing brain networks using noninvasive electrophysiological measurements (EEG/MEG) based on arXiv:2607.17602v1 'Exploring Brain Networks Using Noninvasive Electrophysiological Measurements: Methods and Applications'
Neuropsychology-inspired EEG-to-Text benchmark (COFETT) that addresses EEG instability and enables teacher-forcing-free evaluation, providing evidence for real-world EEG2Text feasibility. Use when evaluating or building non-invasive brain-to-text decoders,…
Apply persistent homology to MEA recordings of spontaneous brain-organoid activity. Detects H1 loops and H2 voids in correlation networks, compares them to rate-preserving null models, and identifies non-redundant loop-carrying cores.
Skill for understanding and applying the research from arXiv:2607.13644 "Evaluating Encoding Strategies for Closed-Loop Classification in Biological Neural Networks"
Analysis of Drosophila FC2 circuit mechanism showing that goal maintenance uses normalization rather than winner-take-all selection, with global inhibition from FB5A neurons keeping a single clean activity bump rather than actively choosing between competing…
Grounding Evidence-Aware Reward for Long-Context Reasoning to Reduce Repetitive Copying
The Giant Hippocampus: From Structural Monoculture to a System of Systems - bridging AI architecture design with neuroscientific understanding of brain structure diversity.
Methodology for identifying and interpreting internal mental workspace in language models using Jacobian lens technique, inspired by neuroscience's Global Workspace Theory.
Skill for understanding grounded world models in biological organisms and their implications for future embodied AI, based on arXiv:2607.13560v1.
Grounded world models in biological organisms and future embodied AI - arXiv:2607.13560 - A framework for understanding how biological intelligence builds grounded world models through interaction, with implications for embodied AI
Methodology for analyzing internal fluctuations in Growing Neural Cellular Automata (GNCA) to understand self-maintenance and self-repair mechanisms. Based on arXiv:2607.12403v1.
How Canada uses Claude: Findings from the Anthropic Economic Index
First-passage-time analysis of inter-spike interval (ISI) statistics for excitatory-inhibitory (EI) integrate-and-fire neurons with depolarizing and hyperpolarizing adaptive thresholds. Use when studying stochastic neuronal firing, ISI variability, adaptive…
Skill for understanding and applying the research from arXiv:2607.14086 "Leveraging unlabelled data for generalizable neural population decoding"
Skill for understanding and applying the MOJO framework for leveraging unlabeled data in neural population decoding via masked autoencoding and joint supervised learning.
Measuring LLMs’ impact on N-day exploits
Model human perception, cognition, and decision dynamics as a modular perception-cognition-decision pipeline state-space model. Provides mathematical formulation, stability conditions, and application to rehabilitation control. Use when you need interpretable…
A modular state-space model for human perception, cognition, and decision dynamics that links sensory inputs to behavior through latent internal states while maintaining interpretable connections to neuro-cognitive mechanisms.
Skill for exploring brain networks using noninvasive electrophysiological measurements (EEG/MEG) based on arXiv:2607.17602v1.