<ul><li>Large language models (LLMs) are effective in decision-making due to their vast knowledge but lack reasoning abilities and adaptability.</li><li>A new approach called Causal-aware LLMs integrates the structural causal model (SCM) into decision-making for better learning, adapting, and acting in complex tasks.</li><li>In the learning stage, the LLM extracts causal entities and relations to create a structured causal model of the environment.</li><li>Experimental results in the open-world game 'Crafter' demonstrate the effectiveness of Causal-aware LLMs in achieving a more accurate understanding of the environment and making efficient decisions.</li></ul>

Causal-aware Large Language Models: Enhancing Decision-Making Through Learning, Adapting and Acting

Discover more