Evaluating LLMs' Inherent Multi-hop Reasoning Ability

Open in new window