interpretable-embeddings. Interpretable text embeddings by asking LLMs yes/no questions (NeurIPS 2024)