ML Bridge Glossary

The entry point for external applications to interact with ML Bridge services through REST and WebSocket endpoints.

Participants who contribute computational resources to execute ML tasks on the network, earning rewards for processing and verification.

Agreement among multiple compute providers about the correctness of a task result, achieved through voting and comparison mechanisms.

Smart contract implementing result verification logic through multi-provider consensus mechanisms.

Mathematical evidence that verifies the correctness of computations without revealing sensitive information.

Consensus-Verified Computation Protocol - ML Bridge's novel mechanism ensuring result quality through multi-provider verification and economic incentives.

Security achieved through financial incentives and penalties rather than trust, ensuring honest behavior through economic mechanisms.

Transaction costs paid in ETH to execute operations on the Ethereum blockchain, including ML Bridge interactions.

The process of using a trained machine learning model to make predictions or generate outputs from new input data.

InterPlanetary File System - a distributed storage protocol used to store AI models and data in a decentralized manner.

The time between task submission and result delivery, typically measured in seconds or minutes.

Scaling solutions like Arbitrum and Polygon that reduce costs and increase throughput for ML Bridge operations.

Descriptive information about AI models including capabilities, requirements, pricing, and performance metrics.

A decentralized platform connecting AI model providers, compute resources, and task requesters through blockchain technology for trustless machine learning task execution.

The native utility token of ML Bridge, used for payments, staking, and rewards.

The core smart contract that orchestrates task submission, provider selection, and result verification across the entire platform.

AI researchers and companies who deploy trained machine learning models on the platform, earning rewards based on usage and performance.

The process of teaching a machine learning algorithm using data to create a model capable of making predictions.

Smart contract managing AI model registration, metadata, versioning, and availability across the network.

The practice of assigning the same task to multiple compute providers to enable verification and prevent single points of failure.

Security mechanism requiring multiple parties to approve transactions, used for critical ML Bridge operations.

The phenomenon where ML Bridge becomes more valuable as more participants (providers and requesters) join the network.

A computing system inspired by biological neural networks, consisting of interconnected nodes that process information.

A compute provider's software installation that connects to the ML Bridge network to receive and process tasks.

The process of validating computation outputs through consensus among multiple compute providers before accepting results.

Smart contract managing token rewards, penalties, and staking mechanisms for network participants.

The penalty mechanism where staked tokens are confiscated for malicious behavior or poor performance.

Self-executing contracts with terms directly written into code, running on the blockchain to automate ML Bridge operations.

The process of locking MLB tokens to participate as a compute provider, demonstrating commitment and enabling penalty enforcement.

A machine learning job submitted to the network, containing input data, model selection criteria, and payment for execution.

Individuals or organizations who submit machine learning tasks to the ML Bridge network, paying for computation and receiving verified results.

Smart contract handling task creation, assignment to compute providers, and lifecycle management.

The number of tasks the ML Bridge network can process per unit time, measured in tasks per hour.

The percentage of time the ML Bridge network is operational and available for task processing.

The practice of earning additional MLB tokens by providing liquidity or participating in specific platform activities.