The post suggests that community members can directly ask their questions on the documentation chat bot. The comments discuss issues with accessing document embeddings in the GPT Index library. Community members tried to print the embeddings, but encountered issues - the doc.embedding attribute returned None even though the embeddings were present in the vector_store.json file. A community member suggested exploring the index._vector_store to look up the embeddings directly. Another community member provided a specific way to access the embeddings using index._vector_store.__dict__["_data"].embedding_dict[doc_id]. However, there is no explicitly marked answer on how to query the index using embeddings.
