Neuron
A neuron is a registered participant in a Bittensor subnet (Glossary: Neuron, Understanding Neurons). The term appears when describing miners, validators, and metagraph state for subnet participants.
Subnet Participant
The term neuron is used inside subnet context. A subnet can contain miner and validator participants, and those participants are represented in subnet state rather than as generic network accounts (Understanding Neurons).
Metagraph Context
The subnet metagraph documentation describes the metagraph as a snapshot containing information about neurons in a selected subnet (The Subnet Metagraph, Glossary: Metagraph). A neuron is therefore one of the participant records whose state can appear in subnet-level views.
Development Stage Context
Subnet development can move from local testing to testchain and then mainchain (Introduction to Bittensor, Bittensor Networks). A neuron observed in one environment should not be read as identical evidence about participant state in another.
For readers, a neuron record belongs to the selected subnet, network environment, and block context that produced the metagraph snapshot (The Subnet Metagraph).
UID and Hotkey-Coldkey Pair
Each neuron in a subnet is identified by a unique UID and linked to a hotkey-coldkey pair for authentication and operations. The official Understanding Neurons documentation describes a neuron as a registered subnet participant—miner or validator—with that UID and key pairing attached.
For readers, neuron is therefore a subnet participant record, not a wallet file label on its own. The UID marks the participant’s slot inside one subnet, while the hotkey-coldkey pair names the keys used to operate that participant on chain.
References: Understanding Neurons, Glossary: Neuron, Glossary: Hotkey-Coldkey Pair
Axon and Dendrite Roles
Official neuron architecture documentation places miners and validators on opposite sides of the same communication model. Miners deploy axon servers to receive validator requests, while validators use dendrite clients to transmit data to those axon endpoints. Synapse objects carry the structured information exchanged through that path.
That server-client split helps readers separate participant role from transport vocabulary. A neuron can be a miner or a validator in subnet terms, while axon, dendrite, and synapse name how subnet messages move between them.
For readers, those transport terms should stay attached to communication context. Axon identifies the miner-side server interface, dendrite identifies the validator-side client interface, and synapse identifies the structured message object exchanged between them.
References: Understanding Neurons, Glossary: Axon, Glossary: Dendrite, Glossary: Synapse
Immunity After Registration
The same neuron lifecycle documentation lists registration and UID assignment as the first steps, followed by an immunity period during which a newly registered neuron is protected from ordinary pruning. Performance metrics such as rank, trust, consensus, and incentive build after that grace window rather than at the instant of registration.
Readers should not treat a fresh UID as immediately exposed to every deregistration rule. Immunity gives a newly registered neuron time to begin producing and being evaluated before low-performance replacement rules apply in the usual way.
References: Understanding Neurons, Glossary: Immunity Period
Relationship to Multiple Mechanisms
A neuron is usually discussed for one incentive mechanism at a time, but subnets can also run multiple mechanisms in parallel. The Glossary notes that validators must evaluate miners separately for each mechanism.
For readers, a neuron still names one registered subnet participant, but the mechanism context matters. Evaluation and performance readings for one mechanism should not be assumed to describe the same neuron in another when a subnet uses more than one evaluation system.
References: Multiple Incentive Mechanisms Within Subnets, Glossary: Multiple Incentive Mechanisms
Relationship to Netuid
A neuron and a netuid are related but different parts of Bittensor subnet vocabulary. A netuid selects which subnet context is in scope, while a neuron names one registered participant record inside that selected subnet. The Glossary: Netuid places netuid at the subnet level, and the Glossary: Neuron describes a registered subnet participant.
For readers, netuid answers which subnet is being discussed, while a neuron answers which participant record exists inside that subnet context.
References: Glossary: Netuid, Glossary: Neuron
Relationship to UID Slot
A neuron and a UID slot are related but different parts of subnet participation vocabulary. A UID slot describes a participant position inside a subnet, while a neuron names the registered participant associated with that position in subnet state. The Glossary: UID Slot describes a participant position, and the Glossary: Neuron describes a registered subnet participant.
For readers, a UID slot names the position identifier, while a neuron names the participant record that occupies that slot in metagraph state.
References: Glossary: UID Slot, Glossary: Neuron
Relationship to Subnet Miner
A neuron and a subnet miner are related but different parts of subnet participant vocabulary. A neuron names a registered participant in subnet state, while a subnet miner names the task-performing role that participant may play inside a subnet. The Glossary: Neuron describes a registered subnet participant, and the Glossary: Subnet Miner describes the entity that performs tasks given by subnet validators.
For readers, a neuron is the participant record term, while a subnet miner is the role term for work production inside a subnet.
References: Glossary: Neuron, Glossary: Subnet Miner
Relationship to Subnet Validator
A neuron and a subnet validator are related but different parts of subnet participant vocabulary. A neuron names a registered participant in subnet state, while a subnet validator names the evaluating role that participant may play inside a subnet. The Glossary: Neuron describes a registered subnet participant, and the Glossary: Subnet Validator describes the role that evaluates miner performance.
For readers, a neuron is the participant record term, while a subnet validator is the role term for evaluation inside a subnet.
References: Glossary: Neuron, Glossary: Subnet Validator
Relationship to Subnet Task
A neuron and a subnet task are related but different parts of subnet design vocabulary. A neuron names a registered participant in a subnet, while a subnet task names the work miners are expected to perform inside that subnet. The Glossary: Neuron describes a registered subnet participant, and the Glossary: Subnet Task describes the work miners perform within a subnet’s incentive design.
For readers, a neuron names a registered participant slot, while a subnet task names the work that subnet asks participants to produce.
References: Glossary: Neuron, Glossary: Subnet Task
Relationship to Incentive Mechanism
A neuron and an incentive mechanism are related but different parts of subnet design vocabulary. A neuron names a registered participant in a subnet, while an incentive mechanism covers the task, protocol, and evaluation method that define how that subnet connects work to rewards. The Glossary: Neuron describes a registered subnet participant, and the Glossary: Incentive Mechanism describes how work and evaluation connect to incentives inside a subnet.
For readers, a neuron names who is registered in a subnet, while an incentive mechanism names the evaluation rules that govern how that participant’s role is judged.
References: Glossary: Neuron, Glossary: Incentive Mechanism
Relationship to Yuma Consensus
A neuron and Yuma Consensus are related but different parts of Bittensor subnet incentive flow. A neuron names a registered participant in a subnet, while Yuma Consensus aggregates validator weight signals from those participants into miner incentives and validator dividends after evaluation. The Glossary: Neuron describes a registered subnet participant, and Yuma Consensus describes clipping, bonding, and emission calculation.
For readers, a neuron names a participant record inside a subnet, while Yuma Consensus names the aggregation process that uses validator submissions from those participants.
References: Glossary: Neuron, Yuma Consensus
Relationship to Validator Weights
A neuron and validator weights are related but different parts of subnet evaluation structure. A neuron names a registered participant in a subnet, while validator weights are the evaluation signals that validator-role participants submit for miners after judging their work. The Glossary: Neuron describes a registered subnet participant, and the Glossary: Validator Weights describes the per-validator evaluation inputs used by Yuma Consensus.
For readers, a neuron names a participant record, while validator weights name the evaluation submissions produced by validator-role participants inside consensus.
References: Glossary: Neuron, Glossary: Validator Weights
Relationship to Subnet Deregistration
A neuron and subnet deregistration describe different levels of state. A neuron is one registered participant inside a subnet, while subnet deregistration removes an entire subnet context (Glossary: Neuron, Subnet Deregistration).
For readers, participant-level deregistration can vacate one neuron slot while the subnet continues. Subnet deregistration ends the subnet context that held all of its neuron records.
Relationship to Child Hotkeys
A neuron and a child hotkey describe different parts of subnet participation. A neuron is a registered participant record, while a child hotkey is a separate key that a parent hotkey authorizes for validation work (Glossary: Neuron, Child Hotkeys).
For readers, the neuron is the participant record. The child hotkey is delegated validation authority, not the participant record itself.
Reader Boundary
This page defines the term at a high level. It does not report current subnet membership, live neuron counts, wallet ownership, or current rewards. Those values change with chain state and should be checked from current Bittensor sources when needed.
Relationship to Coinbase
A neuron and coinbase are related through rewards, but they name different things. A neuron is the registered subnet participant, while coinbase is the per-block emission operation that can trigger epoch processing and reward distribution (Glossary: Neuron, Coinbase Implementation).
For readers, neuron names who can receive miner incentives or validator dividends. Coinbase names part of the mechanism that produces and routes those emissions.
Register Assigns the UID Before Metrics Mature
Understanding Neurons lists registration and UID assignment as the first lifecycle steps on a subnet such as netuid 1, followed by an immunity period before ordinary displacement rules apply in the usual way. Performance fields such as rank, trust, and incentive build after that entry window rather than at the instant of registration.
Neuron vocabulary therefore names a participant record that exists once register succeeds, even before later evaluation metrics reflect sustained work.
Metagraph Exposes One Row Per Neuron on netuid 1
The Subnet Metagraph documentation describes comprehensive neuron information for a selected subnet snapshot. Each registered neuron appears as state inside that metagraph view rather than as a standalone wallet label.
Reading neuron count or participant fields therefore requires both the selected netuid and the block height of the snapshot. A neuron row describes chain-recorded subnet participation at one point in history.