WorldSense benchmark for grounded reasoning in language models - View it on GitHub