Do Large Language Models know who did what to whom?

Re-assign community