CortexFlow Documentation

CortexFlow is a cognitive-inspired memory optimization system for LLMs that implements a multi-tier architecture designed to maximize context utilization and enable complex reasoning.

Key Features

• Multi-tier memory management optimizes token usage

• Importance-based information retention

• Progressive context compression

• Knowledge store for long-term information persistence

• Dynamic memory tier weighting

• Vector-based knowledge retrieval

• Advanced retrieval techniques with GraphRAG

• Chain of Agents framework for complex queries

• Self-reflection for verifying knowledge relevance

• Vertex AI and Ollama backend support

User Guide

• Installation
  • Prerequisites
  • Basic Installation
  • Installation with Optional Dependencies
  • Development Installation
  • Verifying Installation
• Getting Started
  • Basic Usage
  • Using the ConfigBuilder
  • Advanced Features
• Configuration
  • Nested Configuration
  • Builder Pattern
  • Memory Tier Configuration
  • LLM Configuration (Ollama)
  • Vertex AI Configuration
  • Advanced Features Configuration
  • Persistence Configuration
  • Loading from a Dictionary
• Migration from AdaptiveContext
  • Simple Renaming
  • Class Name Changes
  • Temporary Compatibility Layer
  • Configuration Changes
  • Database Compatibility

API Reference

• CortexFlowManager
  • CortexFlowManager
  • Core Methods
  • Example Usage
• CortexFlowConfig
  • CortexFlowConfig
  • Sub-Configuration Classes
  • ConfigBuilder
  • Example Usage

Examples

• Basic Usage

Indices and tables

• Index

• Module Index

• Search Page