Theory of Mind (ToM) was introduced as the capacity to infer the mental states of others, such as desires, beliefs, and intentions. Numerous scenarios involving human cognition and social intelligence rely on the ToM modeling of others’ mental states, such as forecasting others’ actions, planning, and various forms of reasoning and decision-making. However, existing research demonstrates that there is a huge gap in ToM ability between humans and Large Language Models (LLMs), such as ChatGPT and GPT-4. This project focuses on exploring how to evaluate and enhance the ToM ability of LLMs. The potential research methods include building an evaluation dataset and developing state-of-the-art machine learning models for machine mental state reasoning.