How Do Multimodal Large Language Models Handle Complex Multimodal Reasoning? Placing Them in an Extensible Escape Game

show that MLLMs, regardless of scale, can successfully complete the simplest room escape tasks, with some exhibiting human-like exploration strategies. Yet, performance dramatically drops as task difficulty increases. Moreover, we observe that models severely suffer from accidental success, and that performance bottlenecks vary across models, revealing distinct failure modes and limitations in their multimodal reasoning abilities, such as repetitive trajectories without adaptive exploration, getting stuck in corners due to poor visual spatial awareness, and ineffective use of acquired props, such as the key. We hope our work sheds light on new challenges in multimodal reasoning, and uncovers potential improvements in MLLMs capabilities. 1 2