The AGI Manual
Hyperon

Hyperon Overview

Introduction to the OpenCog Hyperon framework for scalable AGI

Hyperon Overview

OpenCog Hyperon is a major redo of the OpenCog AGI framework. It is designed to be faster, more flexible, and more scalable, with a focus on integrating different AI algorithms into a single, unified system.

The Core Vision

Hyperon's goal is to create a "Grand Unified Theory" implementation for AGI, where neural, symbolic, evolutionary, and probabilistic components work together in a shared knowledge space.

Key Components

1. AtomSpace

A distributed, high-performance hypergraph database. It stores "Atoms" (nodes and links) which can represent anything from low-level data to high-level concepts and even program code.

2. MeTTa (Meta Type Talk)

Hyperon's primary programming and query language. MeTTa is:

  • Atom-based: MeTTa code is stored as atoms in the AtomSpace.
  • Reflective: MeTTa programs can reason about and modify themselves.
  • Multi-paradigm: Supports functional, logic, and imperative styles.

3. Distributed AtomSpace (DAS)

Ensures that the knowledge base can scale across many machines, allowing for massive-scale AGI experiments.

4. PLN (Probabilistic Logic Networks)

A reasoning engine that performs uncertain inference over the AtomSpace, combining logic and probability.

Why Hyperon?

  • Interoperability: Allows different AI tools to "talk" to each other through the shared AtomSpace.
  • Neuro-symbolic by Design: Built from the ground up to bridge the symbolic and neural worlds.
  • Self-Modification: The system is designed to eventually be able to rewrite its own code to become more efficient—a key requirement for recursive self-improvement in AGI.

Next: The AtomSpace

On this page