Hyperon is SingularityNET's open-source platform for Artificial General Intelligence, designed to progress from here to AGI and ultimately to beneficial ASI (Artificial Superintelligence). Building on decades of research from the OpenCog project, Hyperon provides a unified neurosymbolic framework where diverse cognitive processes — symbolic reasoning, probabilistic inference, neural learning, evolutionary search, and attention allocation — interoperate over shared memory to produce emergent general intelligence.
Unlike narrow AI systems optimized for individual tasks, Hyperon is designed as a composable infrastructure in which multiple learning and reasoning algorithms collaborate through a principle called cognitive synergy. Each algorithm addresses a fundamental requirement of general intelligence, but it is their interaction — sharing representations, guiding each other's search, and co-evolving within a common knowledge substrate — that is intended to enable the system to tackle problems none could solve alone.
Hyperon's architecture spans four layers, each documented in detail within this wiki:
For decentralized deployment, Hyperon compiles MeTTa into targets running on ASI Chain — a blockchain runtime designed for AGI workloads. ASI Chain provides cryptographically secured execution, content-addressed provenance, and the ability to scale cognitive processes from a single machine to a distributed network. Its F1R3FLY engine renders concurrent process calculi for scalability, while MeTTaCycle orchestrates AGI workloads. The whitepaper describes ASI Chain as targeting "native inference settlement" — verifying cognitive state transitions rather than just token transfers.
Hyperon bridges symbolic and neural paradigms through two complementary approaches:
Hyperon is a ground-up redesign of the earlier OpenCog framework, preserving the core cognitive theories (cognitive synergy, CogPrime architecture, patternist philosophy of mind) while incorporating new ideas at every level: a new type system and language (MeTTa replacing Atomese/Scheme), a new high-performance backend (MORK replacing the C++ AtomSpace), new mathematical controls (quantale-based weakness, geodesic effort), and decentralized execution infrastructure. The transition represents not a departure from OpenCog's vision but its maturation into a system engineered for scalability and composability.
Hyperon inherits its conceptual DNA from OpenCog, an open-source AGI framework that evolved through two decades of research and development. Understanding this lineage illuminates why Hyperon is designed the way it is — and what changed in the transition.
The intellectual roots trace to the late 1990s. Ben Goertzel's work on formalizing general intelligence — defining it as "the ability to achieve complex goals in complex environments" — led to a series of practical systems: the Webmind AI Engine (1997–2001) at Intelligenesis Corp., followed by the Novamente Cognition Engine (2001–2008) at Novamente LLC. In 2008, the Novamente source code was released publicly as OpenCog, establishing an open-source community around the pursuit of AGI.
The conceptual framework was elaborated in several key publications: The Hidden Pattern (2006) on pattern-based philosophy of mind, Probabilistic Logic Networks (2008) on reasoning under uncertainty, and Building Better Minds (2012, with Cassio Pennachin and Nil Geisweiller) detailing the full CogPrime architecture — the specific configuration of cognitive components believed capable of achieving human-level AGI.
The original OpenCog system, now sometimes called "OpenCog Classic," was built in C++, Scheme, and Python around several core components:
These components were deployed in virtual agent control (OpenCogBot in virtual worlds), humanoid robotics (Hanson Robotics integration), and biological knowledge exploration.
Hyperon preserves the core intellectual commitments of OpenCog:
Hyperon is not an incremental update but a ground-up redesign motivated by hard-won lessons from a decade of OpenCog development:
Technical Deep Dive: OpenCog Legacy Full — complete timeline, 11-row architectural bridge map, five motivations for transition, component maturity analysis, CogServer criticisms, anti-CYC philosophy, pattern mining evolution, and maintained vs. archived repos.
Cognitive synergy is the foundational design principle of Hyperon: the idea that general intelligence emerges not from any single algorithm but from the cooperative interaction of multiple specialized cognitive processes sharing a common knowledge substrate. It is both a theory of how minds work and an engineering methodology for building AGI systems.
Human cognition integrates perception, reasoning, memory, attention, motivation, language, and learning into a unified system where each process continuously informs and strengthens the others. A purely logical reasoner struggles with grounding; a purely statistical learner struggles with compositionality; a purely evolutionary system struggles with directed search. But when these approaches operate over shared representations and guide each other's processing, capabilities emerge that none could achieve independently.
In Goertzel's formulation from Building Better Minds: cognitive synergy occurs when multiple cognitive processes — each handling a different aspect of intelligence — cooperate in a way that their combined capability exceeds the sum of their individual contributions. This is not mere parallelism. It is deep interoperation: one process generating hypotheses that another evaluates, one process identifying attention-worthy patterns that another reasons about, one process learning representations that another uses to plan.
Hyperon's architecture is engineered to enable cognitive synergy through three mechanisms:
Some of the most important synergistic interactions designed into Hyperon include:
As discussed in Building Better Minds (Chapter 8), cognitive synergy may explain a puzzling feature of AGI research: the difficulty of measuring partial progress. If intelligence emerges primarily from the interaction between cognitive processes rather than from any individual process, then a system with three out of five components may show dramatically less capability than one with all five — even though it is only "two components away." This creates a perception of sudden capability jumps that are actually the result of crossing synergy thresholds.
This insight has practical consequences for Hyperon development: the system's true capabilities may only become apparent when enough components are integrated and interoperating, not when individual components are benchmarked in isolation.
AtomSpace is the foundational data structure at the heart of Hyperon — a typed, content-addressed metagraph that serves as both shared memory and control plane for all cognitive processes. It is the substrate in which knowledge, code, inference results, attention values, motivational states, and neural representations all coexist and interoperate.
Status: The AtomSpace concept and Space API abstraction are current — implemented in the reference Hyperon codebase (hyperon-experimental) and in MORK. The reference AtomSpace, MORK, and DAS backends are operational. Neural Spaces are experimental; blockchain-backed Spaces via F1R3FLY are proposed.
An AtomSpace is a generalization of a hypergraph. Where a traditional graph connects pairs of nodes with edges, and a hypergraph allows edges to connect arbitrary sets of nodes, a metagraph goes further: edges (Links) can themselves be members of other edges, and both nodes and links carry typed values. This recursive structure naturally represents the nested, multi-relational knowledge structures required for general intelligence — logical implications, probabilistic dependencies, procedural programs, and attention metadata can all be expressed as atoms within the same graph.
Every atom in AtomSpace is typed. The type system encodes data categories (symbols, numbers, expressions), function types, and logical connectives. The scope and enforcement of type constraints varies across backends — the reference implementation and MORK each handle types somewhat differently. MeTTa programs can introspect and manipulate the type structure itself.
A defining feature of AtomSpace is that programs and data share the same representation. A MeTTa function definition, a PLN inference rule, a MOSES-generated program, and a factual knowledge assertion are all atoms in the metagraph. This homoiconicity enables:
AtomSpace is accessed through a universal Space API that abstracts over multiple possible backends. A Space supports operations for adding, removing, querying, and pattern-matching atoms. Current and planned backends:
Atoms in AtomSpace are content-addressed — identified by what they contain rather than where they are stored. In MORK, this is implemented through content identifiers (CIDs) derived from the atom's path in the trie structure, enabling efficient verification and deduplication. Content addressing is what makes distributed and decentralized AtomSpaces practical: the same atom resolves to the same identifier regardless of which node stores it.
Atoms carry associated values beyond their structural content:
The AtomSpace concept originated in the OpenCog project as a C++/Scheme hypergraph database. The original implementation stored atoms in an in-memory graph with optional PostgreSQL persistence. In Hyperon, the concept has been generalized: AtomSpace is now an abstract interface (the Space API) that can be implemented on different backends — from MORK's prefix trees to DAS's distributed storage — while preserving consistent semantics for cognitive processes built on top of it.
Technical Deep Dive: AtomSpace Full — metagraph formal definition, Atom/Value distinction, TruthValue-to-FloatValue transition, Space API abstraction, classical C++ vs. Rust implementations, performance architecture, and historical design decisions.
A central challenge of AGI is unifying the complementary strengths of neural networks (pattern recognition, generalization from data, continuous optimization) with symbolic systems (compositional reasoning, interpretability, knowledge representation). Hyperon's design addresses this through a dual-path architecture offering both pragmatic interoperability and deep structural unification.
Status: Outside integration is current (implemented via MeTTa-Motto). Inside integration (QuantiMORK) and the advanced techniques below are proposed — described in the 2025 whitepaper as research directions.
Hyperon provides two complementary modes of neural-symbolic integration, suited to different stages of system maturity:
In outside mode, existing neural models — large language models, vision transformers, embedding models, reinforcement learning agents — are wrapped as Spaces within Hyperon's Space API. Symbolic processes can query neural models through the same interface they use to query the AtomSpace:
The MeTTa-Motto library implements this approach, embedding LLMs (ChatGPT, Claude, open-source models) as programmable MeTTa functions with support for stateless wrappers, stateful dialogue agents, retrieval-augmented generation, and functional calling. (See MeTTa-Motto for details.)
The whitepaper also describes Symbolic Transformer Heads as part of outside integration: mined patterns from AtomSpace serve as structured templates augmenting standard transformer attention heads, using contrastive symbolic alignment to ensure patterns survive memory compression in compressive transformer architectures.
The 2025 whitepaper proposes a more radical inside mode called QuantiMORK, in which neural network structures would be natively encoded within the MORK metagraph rather than wrapped. This would represent neural network components — weight matrices, activation patterns, wavelet-structured tensors — directly as paths and values in MORK's prefix-tree database.
Proposed properties of QuantiMORK:
To address the risk that neural learning updates could destabilize symbolic knowledge (and vice versa), the whitepaper explores weakness-based regularization: a quantale-valued simplicity metric applying uniformly to both neural weight updates and symbolic inference steps. The whitepaper investigates conditions under which combined updates would approximately commute — a desirable property for reliable integration, though the extent of this commutativity in practice remains an open research question.
A system capable of general intelligence must eventually be capable of reflecting on and improving its own cognitive processes. Hyperon's design treats self-modification as a first-class capability governed by formal mathematical guarantees — aiming to ensure that a system improving itself remains aligned with its intended goals and values.
Status: Proposed. The self-modification pipeline described here is a research design from the 2025 whitepaper (§8). It has not yet been implemented end-to-end. The mathematical foundations (weakness theory, geodesic control) are under active development; the deployment pipeline and governance mechanisms remain architectural proposals.
Self-modifying AGI presents a fundamental tension: the same capability that enables a system to improve itself could allow it to alter its goals, remove its safety constraints, or destabilize its own reasoning. Hyperon's proposed approach is to make self-modification transparent, auditable, and mathematically bounded.
The whitepaper describes a disciplined pipeline for self-modification:
The whitepaper proposes addressing goal stability through supermartingale potentials — Lyapunov-like mathematical functions that provably do not increase under permitted modifications. If a proposed change would increase the potential (indicating goal drift), it would be flagged for additional review or rejected. The aim is to transform goal stability from a philosophical concern into a tractable mathematical problem.
The design envisions certain principles enforced globally across all cognitive processes:
When Hyperon operates on ASI Chain, self-modifications could become subject to multi-party governance — requiring approval from multiple stakeholders, community voting, or smart contract constraints encoding organizational policies.
Technical Deep Dive: Self-Modification and Safety Full — typed metamorphism formalism, supermartingale goal stability, five-stage pipeline details, lens laws, drift bounds, and decentralized governance.
Papers: Hyperon for AGI⇒ASI Whitepaper (2025), §9
Status: Active pilot efforts and research workstreams in four domains. Each at early stages with near-term milestones defined in the whitepaper.
Hyperon's development philosophy is to build real applications in diverse domains from day one, using each as both a testbed and a source of learning that benefits all others. Four domains are currently active — game AI, humanoid social robotics, bioinformatics, and mathematical reasoning. All connect to the same Hyperon+PRIMUS substrate, meaning advances in perception, reasoning, pattern mining, motivation, and transfer benefit every domain simultaneously.
Architecturally, multiple applications share common "collective knowledge" AtomSpaces while maintaining separate per-application AtomSpaces for task-specific processing. The same mathematical controls (weakness priors, geodesic f·g control) and the same neurosymbolic loop (Pattern Miner → WILLIAM → Symbolic Heads/PLN) operate uniformly across all domains.
Responsible: Ben Goertzel (architecture); Berick Cook (AIRIS integration)
Papers: Hyperon for AGI⇒ASI Whitepaper (2025), §9.1
GitHub: Vereya (Minecraft mod), minecraft-demo (Python API), minecraft-experiments, AIRIS-client, rocca (Rational OpenCog Controlled Agent — OpenAI Gym RL), axiom (AXIOM object-centric game-world agent)
Status: Active pilot. Minecraft integration operational via Vereya mod and tagilmo Python API. AIRIS demonstrated causal learning in Minecraft without pre-training. Sophiaverse and Neoterics playground are under development.
Games provide ideal sandboxes for testing perception, planning, tool use, dialogue, and social norms with rapid iteration and safe failure. Three environments are targeted:
The whitepaper describes the game interface operating through Spaces that mirror game state (voxels, entities, quests, markets) into AtomSpace as typed Atoms. Action affordances appear as SubRep options (navigate, craft, trade, negotiate) with MeTTa rules expressing pre/post-conditions. External LLMs handle narration and quest generation initially, with Symbolic Heads retrieving mined templates like task frames and recipes. PLN factor-graphs encode precondition networks and multi-step plans, with geodesic control scoring candidate steps.
AIRIS (Autonomous Intelligent Reinforcement Inferred Symbolism) has demonstrated causal learning in Minecraft — constructing deterministic environment models through direct interaction without pre-training, and self-correcting via scientific method application.
Responsible: Ben Goertzel (architecture)
Partners: Mind Children (education robot); Hanson Robotics (expressive humanoids — Sophia, Desdemona)
Papers: Hyperon for AGI⇒ASI Whitepaper (2025), §9.2
GitHub (legacy OpenCog): ghost_bridge (ROS-GHOST bridge), ros-behavior-scripting (Eva robot control, production 2015–2017), loving-ai (Loving AI dialogue/meditation), loving-ai-ghost (GHOST chatbot scripts for Loving AI)
Status: Partnerships established with Mind Children and Hanson Robotics. Legacy OpenCog ROS integration exists (ghost_bridge, ros-behavior-scripting, loving-ai). Hyperon-native robotics integration is under development; the whitepaper (§9.2) describes the target architecture.
Humanoid social robotics provides a stress-test for PRIMUS: robots must combine perception, dialogue, motor control, and social norms seamlessly — from classroom tutoring to stage performance — without requiring a complete rewrite for each scenario.
The whitepaper describes a layered architecture: external audio/vision neural networks for initial perception, with progressive embedding of predictive coding layers for latency-critical paths (gaze estimation, lip-sync, gesture segmentation). Pattern Miner would discover conversational templates (adjacency pairs, tutoring frames) for Symbolic Heads to retrieve during speech decoding, while PLN enforces persona and safety constraints. MetaMo would budget the balance between exploration and de-escalation behaviors.
Motor control keeps low-level servo loops on the robot while SubRep options cover higher-level skills ("look at face", "nod", "hand-wave") with formal admission certificates. Geodesic control aligns dialogue steps with motion options so physical gestures support conversational goals. The whitepaper describes capability-gated on-device modification and safety dashboards tracking invariant bands for persona and etiquette.
The existing repo surface is legacy OpenCog infrastructure:
Hyperon-native equivalents of these components are part of the development roadmap.
Responsible: Ben Goertzel (architecture); iCog Labs (data pipelines)
Papers: Hyperon for AGI⇒ASI Whitepaper (2025), §9.3
GitHub: agi-bio (genomics/proteomics), biochatter-metta (NL-to-MeTTa queries), pubchem2metta (PubChem conversion), bio-semantic-parser (biological data parsing), bio-data-semantic-parsing (bio data semantic parsing pipeline)
Status: Active pilot. Data ingestion pipelines operational (agi-bio, biochatter-metta, pubchem2metta). End-to-end hypothesis generation on longevity datasets is a near-term milestone.
Biology is fundamentally structured as graphs: genes connect to proteins, proteins form pathways, pathways influence phenotypes, drugs modulate these relationships. Hyperon's graph-native architecture is well-suited to combining noisy biological graphs, mining meaningful motifs, running uncertain chains of reasoning, and proposing ranked hypotheses with clear rationales.
Data flows into AtomSpace through BioSpace adapters that transform omics matrices into node attributes, protein-protein interactions and pathways into edges, literature triples into assertions with provenance, and clinical outcomes into noisy links with confidence scores. Everything receives CIDs, making merges auditable and reproducible. Existing tools include:
The whitepaper describes a pipeline where Pattern Miner identifies motifs (e.g., "gene A ↔ pathway P ↔ phenotype Y with drug D evidence") ranked by I-surprisingness, WILLIAM promotes frequent subgraphs to reusable templates, PLN factor-graphs propagate graded truth over ontologies and experimental results, and MOSES/GEO-EVO evolves predictive programs. TransWeave would move mechanism components across cohorts or omics platforms when matches hold strong.
The proposed output would be ranked hypothesis packs — auditable CID bundles containing mechanism graphs, predictors, expected biomarkers, and counter-evidence — with experiment selection guided by geodesic f·g control.
Responsible: Ben Goertzel (architecture)
Papers: Hyperon for AGI⇒ASI Whitepaper (2025), §9.4
GitHub: chaining (forward/backward chaining, proof synthesis), mm2metta (Metamath→MeTTa converter), mmverify.py (Metamath verifier)
Status: Existing tools include MeTTa forward/backward chaining (chaining repo), Metamath-to-MeTTa conversion (mm2metta), and a Metamath proof verifier (mmverify.py). The whitepaper reports an MM2 proof verifier implemented inside MORK as a fast graph-local proof kernel. The broader automated conjecturing pipeline described below is the proposed research direction.
Most automated theorem proving focuses on proving stated goals, but mathematicians spend much of their time on conjectures — formulating new definitions, lemmas, and "what if" hypotheses that later get proved, refuted, or refined. Hyperon targets automated conjecturing as the primary goal, then uses formal proof for validation.
The whitepaper describes a pipeline built around the MORK/MM2 proof verifier kernel:
Peer-reviewed publications, books, and foundational research underpinning the Hyperon framework. This section provides a browsable index organized by era and topic — from the current Hyperon-era papers to the legacy OpenCog research corpus. Individual component cards throughout the wiki also link to their relevant papers directly.
For structured reading guides that map source material to wiki content, see Publication Maps under Reference.
| Bucket | What it means | Examples |
|---|---|---|
| Dedicated book-length publication cards already exist | These titles already have standalone cards under Publications and are the fastest way to orient a reader. | The Hidden Pattern; Probabilistic Logic Networks; Engineering General Intelligence; Real World Reasoning |
| Important Ben Goertzel books already referenced elsewhere, but not yet broken out here | These works are already named in +Papers and should be the next priority if book coverage is expanded. | The Structure of Intelligence; Chaotic Logic; Artificial General Intelligence; OpenCog: A Software Framework for Integrative AGI; The AGI Revolution |
| Bucket | What it means | Representative items |
|---|---|---|
| Standalone publication cards already exist | These works already have dedicated cards under Publications and are linked throughout the sections below. | The Hidden Pattern; Probabilistic Logic Networks; OpenCog Hyperon: A Framework for AGI at the Human Level and Beyond; Meta-MeTTa; MetaMo: A Robust Motivational Framework for Open-Ended AGI |
| Already represented in RawData or Publication Maps, but not yet promoted to standalone publication cards | These papers are already in the wiki's research pipeline and remain good candidates for future publication-card promotion. | A General Theory of General Intelligence; What Kind of Programming Language Best Suits Integrative AGI?; Patterns of Quantum Cognition II: Quantum Logic as a Foundation for AGI |
| Internal or prepublication design-note layer | These items are valuable to technical readers, but they are not all conventional public papers and should be treated separately from the peer-reviewed / arXiv publication set. | Hyperseed-1: Core Ontology; HyperClaw: Cognitive Orchestration via Attention-Metaprotocol |
| Era | Focus | Key Publication |
|---|---|---|
| Philosophical (1993–2006) | Patternist theory of mind | The Hidden Pattern (2006) |
| Experimental (2006–2014) | CogPrime cognitive engine | Engineering General Intelligence (2014) |
| Foundational (2021–2023) | MeTTa language and MORK substrate | OpenCog Hyperon Whitepaper (2023) |
| Applied (2024–2025) | Multi-domain AGI deployment | AGI-25: PRIMUS, MetaMo, PLN/NARS papers |
| Orchestration (2026–) | Hybrid AGI coordination | HyperClaw: Cognitive Orchestration (2026) |
Technical design documents and formal proofs defining core Hyperon components. Items marked [internal] are not yet publicly available.
Architecture & Framework
MeTTa Language & Formal Semantics
Knowledge Representation & AtomSpace
Predictive Coding & Cognitive Architecture
Reasoning (PLN, NARS, NACE)
Motivation & MetaMo
Neural-Symbolic Integration & LLMs
Formal Verification & Mathematics
Evolutionary Learning (MOSES)
Papers and proposals positioning Hyperon capabilities in specific real-world domains.
Reference layers for researchers, editors, and AI agents working with the Hyperon Wiki. These cards provide cross-cutting indexes, repository catalogs, and source-to-card mappings that support the editorial process.
The Hyperon ecosystem encompasses the companies, research labs, and open-source projects building products and services on top of the Hyperon framework, MeTTa language, and the broader ASI Alliance infrastructure. Unlike a traditional software ecosystem organized around a single vendor, Hyperon's ecosystem reflects its origins in the OpenCog project — a decentralized research community where multiple independent organizations contribute to a shared AGI infrastructure.
At the organizational apex sits the ASI Alliance, formed in June 2024 through the merger of SingularityNET, Fetch.ai, and Ocean Protocol (which later withdrew). The Alliance provides the governance umbrella, the unified ASI token, and the ASI Chain blockchain runtime. Beneath this, TrueAGI serves as the dedicated engineering arm developing Hyperon's core — the MeTTa language, MORK kernel, and PLN reasoning — while F1R3FLY provides the formal semantics and blockchain execution layer through Rholang and the rho calculus.
Applied ecosystem partners build domain-specific products on this shared infrastructure: Rejuve in longevity and bioinformatics, Magi in AGI tooling and education, EARTHwise in sustainability knowledge systems, SophiaVerse in gamified AI worlds, and Mind Children and Hanson Robotics in social robotics. iCog Labs provides contract R&D across multiple domains, contributing 49+ repositories. NuNet and Singularity Finance handle decentralized compute and DeFi infrastructure respectively.
Responsible: Greg Meredith
GitHub / Documentation: https://github.com/F1R3FLY-io/MeTTa-Compiler
Description:
MeTTaTron is the F1R3FLY-native MeTTa compiler, providing a path from MeTTa into MeTTa-IL and serving as the MeTTa implementation most closely aligned with the F1R3FLY / ASI Chain execution stack. Within the broader Hyperon ecosystem, it represents an important route by which MeTTa programs can move toward lower-level runtime environments designed for concurrency, distributed execution, and blockchain-native settlement.
Where Hyperon Experimental functions as the reference implementation and PeTTa emphasizes high-performance symbolic execution, MeTTaTron is best understood as a compiler-oriented bridge between MeTTa source programs and the F1R3FLY-side execution model. This makes it especially relevant wherever MeTTa code must interoperate with MeTTa-IL, Rholang-adjacent infrastructure, or ASI Chain-facing runtime components. MeTTaTron ensures that MeTTa programs can compile downward into concrete, scalable, decentralized execution environments.