Semantic Lens - LLM Knowledge Explorer

Semantic Lens is a powerful tool that combines LLM-driven knowledge extraction with interactive visualization. It helps you explore and understand complex knowledge domains by:

1. Using LLMs to generate structured knowledge hierarchies
2. Creating intuitive, interactive visualizations of these knowledge structures

Features

Knowledge Generation

• LLM-Powered Extraction: Leverages advanced language models to build comprehensive knowledge structures
• Hierarchical Organization: Automatically organizes knowledge into meaningful hierarchies
• Rich Content: Generates detailed explanations and relationships between concepts
• Extensible Framework: Supports different LLM providers (OpenAI, Anthropic, etc.)

Interactive Visualization

• Circular Layout: Intuitive display of hierarchical relationships
• Dynamic Navigation: Click through different levels to explore sub-domains
• Visual Grouping: Clear visual distinction between different knowledge levels
• Responsive Design: Adapts to different screen sizes and devices

Levels

Level 0: Fundamental Worldview/Framework

• The philosophical lens through which all knowledge is viewed
• Establishes the core organizing principles and categories

Level 1: Major Components

• How each philosophical category manifests
• Breaks down each framework category into its primary aspects

Level 2: Specific Domains

• The actual areas of study/knowledge that exemplify these components
• Maps abstract principles to concrete fields

Level 3: Subfields and Specializations

• The concrete disciplines and areas of focus
• Specific established fields of study

Level 4: Topics

• Specific areas within each discipline
• Core concepts and areas of investigation

Test Outputs from Key LLMs

Sonar-small-128k-chat (Perplexity) - All Levels

• Knowledge Map
• Data Output

Gemini 2.0 Flash - Levels 0 and 1

• Knowledge Map
• Data Output

GPT-4o - Levels 0 and 1

• Knowledge Map
• Data Output

Sonnet 3.5 v2 - Levels 0 and 1

• Knowledge Map
• Data Output