Describe the bug
When using the Python implementation of the SemanticTextMemory and calling the save_information_async function the created MemoryRecord contains an embedding that is an array of arrays (functionally a list of vectors) that only contains 1 vector. This forces the need to decompose the list to get a single array (vector) which is all that you should have since only one piece of text is sent for embedding.

To Reproduce
Steps to reproduce the behavior:

1. Go to the semantic_text_memory.py file and place a breakpoint in the save_information_async functionon line 68 shown below
   await self._storage.upsert_async(collection_name=collection, record=data)
2. Run the following test file in debug mode -

import asyncio
import semantic_kernel as sk
import semantic_kernel.connectors.ai.open_ai as sk_oai
import semantic_kernel.memory.volatile_memory_store as sk_mv


async def test():
    # Create a new kernel
    kernel = sk.Kernel()

    # Regiter openai embedding service
    api_key, org_id = sk.openai_settings_from_dot_env()
    kernel.add_text_embedding_generation_service(
        "ada", sk_oai.OpenAITextEmbedding("text-embedding-ada-002", api_key, org_id)
    )

    # Create a new memory store
    kernel.register_memory_store(sk_mv.VolatileMemoryStore())

    await kernel.memory.save_information_async("test", id="test1", text="hello world")


if __name__ == "__main__":
    asyncio.run(test())

4. Inspect the embedding property of the created MemoryRecord
5. See in that the embedding is an array of arrays (Note the same behaviour occurs in save_reference_async and search_async when creating the embedding)

Expected behavior
The embedding passed on to the memory store should be a single array (a vector) instead of an array of arrays (array of vectors) since these functions intend on working against a single input (thus will only create 1 embedding) and call down memory store functions that are also intended to work against a single input embedding.

Screenshots
If applicable, add screenshots to help explain your problem.

Desktop (please complete the following information):

• OS: Windows 11
• IDE: VSCode-Insiders
• NuGet Package Version: N/A
• Python Package/Branch: SK/Main

Additional context
Just as a note, I believe that the generation of a list from the OpenAITextEmbedding is valid as the service supports batching, it is just in the case of these functions they do not decompose the top level list.