Investigating Video Reasoning Capability of Large Language Models with Tropes in Movies

Open in new window