hi i was testing the semantic search using vector search embedding. from the docs the recommend step to query is

typesense_results = typesense_client.multi_search.perform({
    "searches": [{
        "q": "*",
        "collection": COLLECTION_NAME,
        "vector_query": f"content_vector:([{','.join(str(v) for v in embedded_query)}], k:{4})"
    }]
}, {})

i was wondering how in typo tolerance of typesense, can we use multiple embedding columns and apply query_by_weights and sort_by if i want to priotize two different embedding columns.