Last Update
Last update is a per-neuron value in a subnet’s metagraph that records the block at which a neuron’s data was most recently updated. Official documentation presents it as the marker of a neuron’s latest update and ties it directly to whether that neuron counts as currently active.
References: The Subnet Metagraph
What It Records
The value is a block height, not a wall-clock time. It captures the most recent point at which a neuron’s state changed on the subnet, such as a validator setting weights. Because it is measured in blocks, its meaning is read against the current block rather than against a calendar.
Reference: The Subnet Metagraph
Determining Activity
Last update is the input to a subnet’s activity test. The documentation describes a neuron as inactive once its last update is far enough in the past: when the last update plus the activity cutoff is below the current block, the neuron is treated as inactive. Until that point it is still within the active window.
References: The Subnet Metagraph, Subnet Hyperparameters
Relationship to Activity Cutoff
Last update supplies the “when,” and the activity cutoff supplies the “how long.” On their own neither decides activity; together they do. The cutoff sets how many blocks a neuron may go without updating, and last update is the timestamp the cutoff is measured from, so the two values are always read as a pair.
References: The Subnet Metagraph, Subnet Hyperparameters
Why It Matters
The point of tracking last update is to keep consensus based on participants that are currently contributing. A neuron whose last update has fallen outside the activity window is treated as inactive, so a stale or absent participant does not keep influencing an epoch’s outcome. This is how the network distinguishes recently active neurons from ones that have gone quiet.
Reference: The Subnet Metagraph
Development Stage Context
The Introduction to Bittensor describes subnet development as moving from localnet to testnet and then mainnet. For last update, that sequence changes how readers should interpret neuron activity timestamps and metagraph snapshot examples.
In localnet, last-update fields can be tested in an isolated environment. Localnet block heights do not represent production neuron activity.
On testnet, metagraph activity fields can be exercised in a shared non-production network. Testnet last-update values are separate from mainnet subnet state.
On mainnet, last update describes live production neuron activity on subnets such as netuid 1. Observed timestamps depend on the selected subnet, block height, and activity-cutoff settings (The Subnet Metagraph).
The Bittensor Networks reference separates mainnet, testnet, and localnet. A last-update example from one environment should not be read as representing production neuron activity in another environment.
Relationship to Yuma Consensus
Last Update and Yuma Consensus describe related parts of Bittensor’s incentive system. Yuma Consensus is the on-chain process that aggregates validator weight signals within a subnet into miner incentives and validator dividends, applying consensus clipping, bonding, and emission calculation (Yuma Consensus).
For readers, last update names a specific part of that incentive picture, while Yuma Consensus names the consensus process that turns validator weights into the resulting incentives and dividends.
Reader Boundary
This page defines the concept at a high level. It does not report the last update of any particular neuron, whether a given neuron is active right now, or the activity-cutoff value on a subnet. Those are live chain state and should be checked for the relevant netuid, for example netuid 1. The activity rule described here is the documented relationship, not a live reading.
Reference: The Subnet Metagraph
Metagraph Views Expose Last Update Per Neuron
The Subnet Metagraph presents subnet neurons as structured rows within a metagraph snapshot. Last update is one of the per-neuron fields readers encounter when inspecting that row data alongside stake, rank, and incentive-related values.
That placement keeps last update in observation vocabulary. It is a field returned with other neuron state in a subnet snapshot, not a separate wallet label or account setting.
References: The Subnet Metagraph, Glossary: Metagraph
Snapshots Are Tied to a Block Height
Subnet metagraph documentation describes the metagraph as a snapshot of subnet state at a particular block. Last update uses the same block- height vocabulary: it records the block of a neuron’s most recent change, which readers then compare against the snapshot’s current block context.
Because both values are block-based, the field is meant to be read inside chain timing rather than as a calendar timestamp.
References: The Subnet Metagraph, Inspecting the Chain with Polkadot.js
Activity Fields Sit Beside Incentive Metrics
Official metagraph material groups last update with other neuron activity and incentive-related values such as ranks, trust, and consensus scores. The activity test that combines last update with activity cutoff therefore sits among the same row of chain-derived performance fields rather than in a separate identity record.
Readers comparing whether a neuron is active should keep that row context in view. Last update is one activity input within a broader metagraph snapshot, not a standalone status badge.
References: The Subnet Metagraph, Understanding Neurons