monad.ai / WeightReport
Type Alias: WeightReport
WeightReport =
object
Defined in: kernel/adaptiveWeights.ts:101
A point-in-time snapshot of learned scorer weights with change context.
Returned by getWeightReport and exposed via GET /.mesh/weights.
Properties
current
current:
Record<string,number>
Defined in: kernel/adaptiveWeights.ts:103
Current learned weights (same keys as DEFAULT_WEIGHTS plus any custom scorers).
defaults
defaults:
Record<string,number>
Defined in: kernel/adaptiveWeights.ts:105
Baseline values the system started from (hard-coded defaults).
delta
delta:
Record<string,number>
Defined in: kernel/adaptiveWeights.ts:107
current - defaults per scorer; positive = reinforced, negative = penalized.
health
health:
WeightHealth
Defined in: kernel/adaptiveWeights.ts:121
Diagnostic health signals for the learning loop. See WeightHealth.
lastUpdatedAt
lastUpdatedAt:
number|null
Defined in: kernel/adaptiveWeights.ts:111
Unix millisecond timestamp of the most recent weight update, or null if never updated.
namespace?
optionalnamespace?:object
Defined in: kernel/adaptiveWeights.ts:123
Namespace-local report when getWeightReport(namespace) is requested.
blended
blended:
Record<string,number>
current
current:
Record<string,number>
delta
delta:
Record<string,number>
maturity
maturity:
number
namespace
namespace:
string
sampleCount
sampleCount:
number
stable
stable:
boolean
Defined in: kernel/adaptiveWeights.ts:119
True when no delta exceeds 5% of its default weight.
A stable system has not yet learned much, or has converged back to near-default weights after a period of learning. Not necessarily a problem — a homogeneous mesh naturally converges to defaults.
updateCount
updateCount:
number
Defined in: kernel/adaptiveWeights.ts:109
Total number of gradient steps applied since the daemon started (or since last reset).