Toolkit to help understand "what lies" in word embeddings. Also benchmarking! - View it on GitHub