Different query results with same query vector but different index type (base64 vs regular)

Steps involved:-

1. Load siftsmall data along with its base64 encodings 
2. Create 2 indexes one vector based and other vector_base64.
3. Run a normal vector query on vector based index

   • { "query":{ "match_none":{ } }, "explain":true, "knn":[{ "field":"vector_data", "k":2, "vector":[1,3,11,110,62,22,4,0,43,21,22,18,6,28,64,9,11,1,0,0,1,40,101,21,20,2,4,2,2,9,18,35,1,1,7,25,108,116,63,2,0,0,11,74,40,101,116,3,33,1,1,11,14,18,116,116,68,12,5,4,2,2,9,102,17,3,10,18,8,15,67,63,15,0,14,116,80,0,2,22,96,37,28,88,43,1,4,18,116,51,5,11,32,14,8,23,44,17,12,9,0,0,19,37,85,18,16,104,22,6,2,26,12,58,67,82,25,12,2,2,25,18,8,2,19,42,48,11]}] }

• • Result Doc Ids
    • vect2177
    • vect3753

1. Run the same query but on base64 encoded out of the same vector

• • { "query": {"match_none": {} }, "explain": true, "knn": [ { "field": "vector_data_base64","k": 2,"vector_base64":"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"} ]

• • Result doc Ids :
    • vect190
    • vect224

To Note: The results for vector type query are not present in top 10 results of base64 encoded query.