Cohorts

Cohort intelligence gives the Twin population context before personalization begins.

DT4H uses cohorts as structured population context, not marketing segmentation. Cohorts define comparable baselines, expected variance, and inference priors.

Cohort diagram #

INPUTNormalized Signals

Biomarkers · behavior · clinical context · self-report

COHORT ENGINEProbabilistic Fit

Weighted membership across relevant population contexts

REFERENCE SELECTIONReference Priors

Expected ranges · variance · trajectories

OUTPUTTwin Initialization

Cohort-aware model starting state

Cohort resolution flow #

01

Population context

Places an individual inside comparable biological, behavioral, and clinical contexts.

02

Probabilistic assignment

Represents cohort fit as weighted membership, not a single fixed segment.

03

Reference priors

Feeds reference-human expectations used for initialization and comparison.

04

Variance modeling

Defines expected range rather than reducing populations to averages.

05

Dynamic reassignment

Allows cohort fit to change as the Twin accumulates evidence.

06

Calibration support

Provides context for evaluating drift, progress, and unexpected outcomes.

Boundary principles #

CohortNot identity

Cohort assignment is probabilistic context, not a fixed label.

InferenceNot diagnosis

Cohort fit does not independently determine clinical condition.

ReferenceNot average person

Reference humans are distributional priors, not generic avatars.

CalibrationDynamic context

Cohort assumptions must update as evidence accumulates.

Implementation notes #

Use probabilistic assignment

Cohort membership should be weighted and revisable, not a hard permanent label.

Keep cohorts separate from identity

Cohorts provide modeling context; they should not be presented as fixed user identity.

Support reassignment

As evidence accumulates, cohort fit should update and influence calibration.

LayerCohort Intelligence
StatusActive Draft
SystemDT4H / StateK / SETPOINT
BoundaryInfrastructure, not diagnosis
System lineageDT4HTwinStateKSETPOINTOutcomesRecalibration
Infrastructure boundaryDT4H models cohorts, Twins, calibration, and runtime state. It does not diagnose, prescribe, or replace licensed clinical judgment.
Document statusInfrastructure draft
Last updatedMay 2026
Applies toDT4H.ai / AvatarK.ai ecosystem