Member Junction
    Preparing search index...

    Interface EmbeddingFeatureStep

    embedding — pull a persisted, version-pinned embedding for an entity and expand it into one numeric feature per dimension (plan §5.2). The embedding is never regenerated at scoring time.

    interface EmbeddingFeatureStep {
        Dims: number;
        EmbeddingModelRef: string;
        Entity: string;
        Id: string;
        Inputs?: string[];
        Kind: "embedding";
        Label?: string;
    }

    Hierarchy (View Summary)

    Index

    Properties

    Dims EmbeddingModelRef Entity Id Inputs? Kind Label?

    Properties

    Dims: number

    Number of embedding dimensions (= number of numeric features produced).

    EmbeddingModelRef: string

    Reference to the embedding model (an MJ: AI Models id/name) — pinned into model Lineage.

    Entity: string

    Entity whose persisted embedding is used.

    Id: string

    Stable id for this step, referenced by downstream steps' Inputs.

    Inputs?: string[]

    Ids of upstream steps that feed this step (DAG edges).

    Kind: "embedding"

    Discriminator selecting the concrete step variant.

    Label?: string

    Optional human-readable label for the visual DAG.

    Settings

    Member Visibility

    On This Page

    Properties