Min_score threshold issues

OpenSearch
1

Versions (relevant - OpenSearch/Dashboard/Server OS/Browser):
Opensearch - 2.18.0

Describe the issue:
The min_score filtering on neural query doesnt seem to work after a certain value.
For example: when I make min_score as 500 or 50000 or even a higher value, the results count gets static. Wanted to understand if this an issue or is there any logic behind this.
Note: the space type used here is dot product.

Configuration:
Neural Query

GET demo-products/_search
{
  "track_total_hits": true, 
  "_source": {
    "includes": [
      "product_id",
      "description",
      "Departments_s",
      "Brand_s"
    ]
  },
  "size": 15,
  "query": {
    "hybrid": {
      "queries": [
        {
          "neural": {
            "description_dense_vector": {
              "query_text": "tile",
              "model_id": "zmqgypQBxSnWxgGRDYIB",
              "min_score": 500000
            }
          }
        }
      ]
    }
  }
}

Be it 500 or 5000 i get around 22000 results always. Only when you start increasing min_score from 1 to 10 or 50, the results gets filtered to certain extent.

Relevant Logs or Screenshots: