Setting default similarity to L2 norm for binary embedding type

Author: ymao1

x-pack/plugin/inference/src/main/java/org/elasticsearch/xpack/inference/services/jinaai/JinaAIService.java

@@ -38,6 +38,7 @@
 import org.elasticsearch.xpack.inference.services.SenderService;
 import org.elasticsearch.xpack.inference.services.ServiceComponents;
 import org.elasticsearch.xpack.inference.services.ServiceUtils;
+import org.elasticsearch.xpack.inference.services.jinaai.embeddings.JinaAIEmbeddingType;
 import org.elasticsearch.xpack.inference.services.jinaai.embeddings.JinaAIEmbeddingsModel;
 import org.elasticsearch.xpack.inference.services.jinaai.embeddings.JinaAIEmbeddingsServiceSettings;
 import org.elasticsearch.xpack.inference.services.jinaai.rerank.JinaAIRerankModel;
@@ -294,7 +295,7 @@ public Model updateModelWithEmbeddingDetails(Model model, int embeddingSize) {
         if (model instanceof JinaAIEmbeddingsModel embeddingsModel) {
             var serviceSettings = embeddingsModel.getServiceSettings();
             var similarityFromModel = serviceSettings.similarity();
-            var similarityToUse = similarityFromModel == null ? defaultSimilarity() : similarityFromModel;
+            var similarityToUse = similarityFromModel == null ? defaultSimilarity(serviceSettings.getEmbeddingType()) : similarityFromModel;
             var maxInputTokens = serviceSettings.maxInputTokens();
 
             var updatedServiceSettings = new JinaAIEmbeddingsServiceSettings(
@@ -323,7 +324,10 @@ public Model updateModelWithEmbeddingDetails(Model model, int embeddingSize) {
      *
      * @return The default similarity.
      */
-    static SimilarityMeasure defaultSimilarity() {
+    static SimilarityMeasure defaultSimilarity(JinaAIEmbeddingType embeddingType) {
+        if (embeddingType == JinaAIEmbeddingType.BINARY || embeddingType == JinaAIEmbeddingType.BIT) {
+            return SimilarityMeasure.L2_NORM;
+        }
         return SimilarityMeasure.DOT_PRODUCT;
     }
 

x-pack/plugin/inference/src/test/java/org/elasticsearch/xpack/inference/services/jinaai/JinaAIServiceTests.java

@@ -996,6 +996,7 @@ private void testUpdateModelWithEmbeddingDetails_Successful(SimilarityMeasure si
 
         try (var service = new JinaAIService(senderFactory, createWithEmptySettings(threadPool))) {
             var embeddingSize = randomNonNegativeInt();
+            var embeddingType = randomFrom(JinaAIEmbeddingType.values());
             var model = JinaAIEmbeddingsModelTests.createModel(
                 randomAlphaOfLength(10),
                 randomAlphaOfLength(10),
@@ -1004,12 +1005,14 @@ private void testUpdateModelWithEmbeddingDetails_Successful(SimilarityMeasure si
                 randomNonNegativeInt(),
                 randomAlphaOfLength(10),
                 similarityMeasure,
-                JinaAIEmbeddingType.FLOAT
+                embeddingType
             );
 
             Model updatedModel = service.updateModelWithEmbeddingDetails(model, embeddingSize);
 
-            SimilarityMeasure expectedSimilarityMeasure = similarityMeasure == null ? JinaAIService.defaultSimilarity() : similarityMeasure;
+            SimilarityMeasure expectedSimilarityMeasure = similarityMeasure == null
+                ? JinaAIService.defaultSimilarity(embeddingType)
+                : similarityMeasure;
             assertEquals(expectedSimilarityMeasure, updatedModel.getServiceSettings().similarity());
             assertEquals(embeddingSize, updatedModel.getServiceSettings().dimensions().intValue());
         }
@@ -1866,8 +1869,13 @@ private void test_Embedding_ChunkedInfer_BatchesCalls(JinaAIEmbeddingsModel mode
         }
     }
 
-    public void testDefaultSimilarity() {
-        assertEquals(SimilarityMeasure.DOT_PRODUCT, JinaAIService.defaultSimilarity());
+    public void testDefaultSimilarity_BinaryEmbedding() {
+        assertEquals(SimilarityMeasure.L2_NORM, JinaAIService.defaultSimilarity(JinaAIEmbeddingType.BINARY));
+        assertEquals(SimilarityMeasure.L2_NORM, JinaAIService.defaultSimilarity(JinaAIEmbeddingType.BIT));
+    }
+
+    public void testDefaultSimilarity_NotBinaryEmbedding() {
+        assertEquals(SimilarityMeasure.DOT_PRODUCT, JinaAIService.defaultSimilarity(JinaAIEmbeddingType.FLOAT));
     }
 
     @SuppressWarnings("checkstyle:LineLength")