Data Science & ML Claude 技能

Letta AI Agent Architect

Builds and manages stateful AI agents with long-term memory and persistent context using the Letta (MemGPT) framework.

Transcription Expert

Streamlines audio and video transcription workflows by providing implementation patterns and comparative analysis for Whisper, Deepgram, and AssemblyAI.

Causal Inference & World Models

Implements advanced causal reasoning, counterfactual analysis, and System 2 deep learning architectures to build robust, intervention-aware AI models.

ASI Polynomial Operads

Implements advanced category theory frameworks to build compositional, causal, and explainable intelligence systems using polynomial functors and operads.

Topos Model Generator

Generates sheaf-theoretic models and formal logic structures using forcing semantics and internal topos languages.

Gödel Machine Self-Evolution

Implements self-improving AI systems using formal verification and evolutionary search to safely enhance agent performance.

Causal Inference Reasoning

Implements interventional reasoning and counterfactual analysis to build robust, System 2 world models for AI.

World Hopping Navigation

Navigates complex mathematical and philosophical possibility spaces using Badiou-inspired ontology and triangle inequality constraints.

Self-Evolving AI Agent Framework

Implements Darwin Gödel Machine patterns to build AI agents that autonomously improve their own code and capabilities through open-ended evolution.

Context Window Expert

Optimizes LLM context usage through advanced token management, semantic chunking, and intelligent prompt compression.

Forward-Forward Learning

Implements Geoffrey Hinton’s Forward-Forward algorithm for efficient, local layer-wise neural network training without backpropagation.

GFlowNet Generative Modeling

Implements Generative Flow Networks to sample diverse, high-reward candidates for molecule design, causal discovery, and combinatorial optimization.

Condensed Mathematics Framework

Unifies topology and algebra using the Scholze-Clausen framework to model condensed sets and liquid vector spaces.

Sheaf Cohomology BCI Consistency

Ensures local-to-global signal consistency in Brain-Computer Interface data using cellular sheaves and Cech cohomology.

Stochastic Resonance Analyzer

Enhances subthreshold signal detection using noise-optimization techniques and Kramers escape rate calculations.

Geohash Coloring & Spatial Indexing

Generates deterministic GF(3) colored identifiers for hierarchical spatial indexing and location-based clustering.

Forward-Forward Learning

Implements Geoffrey Hinton's Forward-Forward algorithm to enable local, layer-wise neural network training without backpropagation.

Turing Chemputer

Coordinates programmable chemical synthesis by executing Turing-complete XDL programs on modular robotic hardware.

Möbius Path Filter

Classifies and filters dependency graph paths using Möbius inversion to optimize proof structures and resolve circular logic.

Ramanujan Expander & Spectral Analysis

Constructs and verifies Ramanujan graphs to ensure optimal spectral expansion and network mixing times.

Gödel Machine Self-Improvement

Implements self-improving AI systems using formal verification and evolutionary search to safely enhance agent performance.

Mathpix OCR & ACSet Extraction

Converts mathematical documents and images into structured LaTeX and ACSet data models using resilient balanced ternary checkpoints.

Color Envelope Preserving

Enforces GF(3) ternary color conservation across data navigation paths to ensure deterministic traversal and structural integrity.

LLM Application Development

Builds sophisticated AI-powered applications using advanced prompt engineering, RAG patterns, and multi-provider LLM integrations.

CatColab Petri Nets

Models concurrent and distributed systems using categorical Petri nets to simulate resource flow and event transitions.

Cognitive Surrogate

Builds, trains, and validates high-fidelity psychological models from interaction patterns to simulate cognitive behavior and intellectual trajectories.

IES Polyglot Analysis Environment

Orchestrates a multi-language environment for advanced social data analysis, media processing, and deterministic generative aesthetics using Julia, Clojure, and Python.

Cognitive Surrogate

Builds, trains, and validates high-fidelity psychological models from interaction patterns to predict cognitive trajectories and generate authentic responses.

IES Topological Environment

Orchestrates polyglot environments for social data analysis and multimedia processing using Clojure, Julia, and DuckDB.

SplitMix Ternary Streams

Extends splittable random number generation with GF(3) balanced ternary streams for parallel triadic systems.