Client 01
A Modular Decentralized AI Platform
Become a subnet validator

Stake → Validate → Secure

Run a subnet validator node to validate incentivized and decentralized AI models.

Hypertensor requires an economic stake before becoming a subnet validator as part of its Proof of Stake (PoS) consensus mechanism, a consensus mechanism that ensures the security and integrity of the network.
The blockchain itself uses NPoS (Nominated Proof of Stake), a variation of PoS. It is designed with the roles of validators and nominators, to maximize chain security. Validators assume the role of producing new blocks, validating blocks, and guaranteeing finality. Nominators can choose to back select validators with their stake. Nominators can approve candidates that they trust and back them with their tokens.

Distro-Style Subnets

Hypertensor subnets are required to be distributed and decentralized.
Using blockchain technology; each subnet validator must have a SHA-256 multihash encoded public key.

Censorship Resistant

Control and decision-making are distributed across a network of participants rather than concentrating it on a single central authority. Therefor, models are unable to be censored by anyone.

Feature 03

Blockchain ↔ AI

The Hypertensor blockchain is a consensus layer for keeping decentralized AI models secure and incentivized. This keeps Hypertensor subnets secure for everyone while validators earn themselves TENSOR in the process. Subnets are verified by on-chain consensus mechanisms powered by blockchain validators.
Feature 03
Proof of [-]
Each subnet must have a consensus mechanism, such as PoI (Proof of Inference), PoT (Proof of Training), etc. This data must be submitted by a randomly chosen validator to the blockchain on each epoch for other validators to agree with or dispute through a fault-proofs mechanism.
#subnet-1 uses PoI (Proof of Inference) of one of its many consensus mechanisms. Proof of Inference is a protocol within the subnet that ensures the correct execution of machine learning (ML) models in Hypertensor subnets. In essence, Proof of Inference is designed to verify that a given validator has performed its inference (i.e., prediction or decision-making) correctly and honestly.
Feature 03
Hypertensor Subnets

Let's build