Package com.tnsai.evaluation.rag.embedding

Class TokenBagEmbeddingFunction

java.lang.Object

com.tnsai.evaluation.rag.embedding.TokenBagEmbeddingFunction

All Implemented Interfaces:

com.tnsai.memory.advanced.EmbeddingFunction

public final class TokenBagEmbeddingFunction
extends Object
implements com.tnsai.memory.advanced.EmbeddingFunction

Deterministic bag-of-tokens embedding used as an offline bench baseline.

Cosine similarity tracks shared tokens. Same-language paraphrase can score, but a Turkish query against an English document shares almost no tokens, so cross-lingual recall stays near zero.

Since:

0.14.0

Field Summary

Fields

Modifier and Type	Field	Description
static final TokenBagEmbeddingFunction	DIM_384

Constructor Summary

Constructors

Constructor	Description
TokenBagEmbeddingFunction(int dimensions)

Method Summary

Modifier and Type	Method	Description
int	dimensions()
float[]	embed(String text)

Methods inherited from class java.lang.Object

equals, getClass, hashCode, notify, notifyAll, toString, wait, wait, wait

Field Details

DIM_384

public static final TokenBagEmbeddingFunction DIM_384

Constructor Details

TokenBagEmbeddingFunction

public TokenBagEmbeddingFunction(int dimensions)

Method Details

dimensions

public int dimensions()

embed

public float[] embed(String text)

Specified by:

embed in interface com.tnsai.memory.advanced.EmbeddingFunction