/LLM_inherent_multi_step_eval

the data of paper "Evaluating LLMs’ Inherent Multi-hop Reasoning Ability"