Glossary
Blockchain Fundamentals
Blockchain: A decentralized, append-only database that maintains a cryptographically secured chain of blocks containing transaction records. Each block is linked to its predecessor through cryptographic hashing, ensuring immutability and transparency across the network.
Smart Contract: Self-executing computer programs on a blockchain that automatically enforce and execute agreements. Think of them as digital vending machines that run exactly as programmed.
L1 (Layer 1): A base blockchain network like Ethereum or Ritual Chain that handles its own transaction validation and security.
On-chain: Data or computations that exist directly on the blockchain’s virtual machine.
Off-chain: Data or computations that exist outside the blockchain’s virtual machine, and are brought on-chain as needed.
Cross-chain: The ability to execute transactions and transfer assets between different blockchain networks while maintaining security and atomicity. This enables interoperability and composability across diverse blockchain ecosystems. See how we enable this in Chain Abstraction.
EOA (Externally Owned Account): A regular blockchain account controlled by a private key, like your personal wallet.
Staking: The process of locking up tokens as collateral to participate in network consensus, validation, or other protocol functions. Staked tokens can be slashed for malicious behavior, creating economic security. Related to our Economic Design.
ZK (Zero-Knowledge) Proof: A method to prove something is true without revealing the underlying information.
Core Protocol Components
EVM: The Ethereum Virtual Machine, a Turing-complete state machine that executes smart contract bytecode in a sandboxed environment. It maintains deterministic execution across all network nodes while managing gas consumption and state transitions. Learn more about our extended version at EVM++.
EVM++: Ritual’s enhanced version of the EVM that adds native support for AI operations, scheduled transactions, and advanced features while maintaining backward compatibility. This enables sophisticated smart contracts with built-in AI capabilities. Learn more in EVM++ Overview.
Symphony Protocol: A consensus protocol that implements dual proof sharding, distributed verification, and optimal sampling to reduce redundant execution. It enables the execute-once-verify-many-times paradigm while maintaining security guarantees. Learn more in Symphony.
Resonance Mechanism: A dynamic fee mechanism that matches compute supply with demand through a market-driven approach. It optimizes resource allocation by considering workload characteristics and node specialization. See details in Resonance.
EOVMT (Execute-Once-Verify-Many-Times): An optimization protocol where complex computations are executed by one node and verified efficiently by others through cryptographic proofs. This reduces network-wide computational redundancy while maintaining security. Learn more in Symphony.
Network Architecture
Node Specialization: A network architecture that allows nodes to optimize for specific types of computation (AI, ZK, TEE) while maintaining network consensus participation. This enables efficient resource allocation and parallel processing of heterogeneous workloads. Learn more in Node Specialization.
Proof Sharding: A technique in Symphony that distributes the verification of computation proofs across different node subsets to improve scalability and efficiency. See our implementation in Symphony.
Distributed Verification System: A component of Symphony that enables efficient verification of complex computations through proof partitioning and committee-based consensus. See our approach in Symphony.
Execution Sidecar: A modular component that handles specialized computation tasks asynchronously from the main chain. Sidecars enable complex operations while maintaining EVM compatibility and security guarantees. Learn more in Execution Sidecars.
Chain Abstraction Layer: A protocol layer that enables seamless interaction between different blockchain networks while maintaining security and consistency guarantees. See details in Chain Abstraction.
AI and Model Infrastructure
Model Inference: The process of using a trained AI model to generate outputs from new inputs. This includes preprocessing inputs, running them through the model’s computational graph, and post-processing outputs. Learn about our implementation in AI Primitives.
Fine-tuning: A machine learning technique that adapts a pre-trained model to a specific task by updating its parameters with new data. This preserves the model’s base knowledge while optimizing for targeted use cases. See our approach in AI Primitives.
Verifiable Model Training: The process of training AI models with cryptographic guarantees of correctness and provenance, using techniques like watermarking and backdoors. Learn more in Verifiable Provenance.
Enshrined Model: An AI model that is natively represented on-chain as a first-class citizen, enabling direct smart contract integration and verifiable execution. Learn more in Enshrined AI Models.
Verification and Privacy
TEE (Trusted Execution Environment): A secure hardware enclave that isolates sensitive computations from the main system. TEEs provide confidentiality and integrity guarantees through hardware-based encryption and attestation. See our implementation in TEE Execution.
ZKML (Zero-Knowledge ML): A system that combines zero-knowledge proofs with machine learning to verify model computations without revealing inputs, weights, or intermediate values. This enables private and verifiable AI execution. See our implementation in ZK Proving and Verification.
OPML (Optimistic-ML): A verification system that assumes AI computations are correct by default but allows for challenge periods. Challengers can dispute results by providing counter-examples, leading to efficient verification of large models. See AI Primitives.
PPML (Probabilistic Proof ML): A verification approach that uses statistical properties and backdoors in ML models to prove training authenticity and detect unauthorized modifications. This provides efficient verification for model provenance. Learn more in Verifiable Provenance.
Modular Computational Integrity: A flexible system that allows developers to choose and combine different verification approaches (ZKML, OPML, PPML) based on their specific needs. See our implementation in Modular Computational Integrity.
Platform Features
Oracle: A bridge between blockchains and external data sources that enables smart contracts to access off-chain information in a secure and verifiable way. Oracles validate and authenticate external data before making it available on-chain. See how Ritual implements oracles in Enshrined Oracles.
Infernet: A decentralized oracle network specifically designed for bringing verified AI computation on-chain. It provides secure model execution, result verification, and cross-chain integration capabilities. See Infernet to Chain.
Agent Launchpad: A platform for deploying and managing autonomous agents with built-in economic incentives and security guarantees. It provides standardized interfaces and safety mechanisms for agent-protocol interactions. See Agents.
Model Registry: A decentralized system for registering, tracking, and managing AI models on-chain. It handles model versioning, ownership, and provenance verification. Learn more in Enshrined AI Models.
Model Marketplace: A decentralized platform for trading AI models with built-in provenance tracking, licensing, and automated revenue distribution. Learn more in Model Marketplace.
Ritual Guardians: A specialized node system that monitors network health and verifies complex computations. Guardians form a security layer complementing traditional validators, particularly for heterogeneous workloads. See Guardians.
Ritual Portals: Cross-chain bridges built on GMP protocols that enable secure access to Ritual’s specialized compute capabilities from any blockchain. Portals handle message verification, state synchronization, and result delivery. Learn more in Ritual to World.
Modular Storage: A flexible storage architecture that supports multiple backend solutions while providing a unified interface for developers. This enables optimization for different data types while maintaining consistency and availability. See Modular Storage.
Provenance: A comprehensive system for tracking and verifying the complete lifecycle of digital assets, including creation, ownership transfers, and modifications. This enables trust and accountability in decentralized systems. See our approach in Verifiable Provenance.