AzureOpenAI On Your Data - the weights of vector and simple in vector_simple_hybrid

Question

I am using Azure OpenAI On Your Data with API version 2024-10-01-preview. In the extra_body>fields_mapping section, I have defined content_fields and vector_fields. The field key is included only in content_fields and does not have a vectorized representation in vector_fields. The key field is defined as searchable.

"fields_mapping": {
    "content_fields": [
        "content",
        "key",
    ],
    "vector_fields": [
        "content_vector"
    ]
},

When I make a request of type simple where the prompt contains only a key value, the response includes documents that contain this value in the key field, which is the expected behavior.

However, when I make a request of type vector_simple_hybrid, I need to increase the top parameter to ~30 so that it finds the relevant documents with this key value. I haven't found any documentation on this but how are results from the vector and simple search weighted and can this be configured?