1 paper
Xue Liu, Xin Ma, Yuxin Ma +36
As Large Language Models (LLMs) exhibit plateauing performance on conventional benchmarks, a pivotal challenge persists: evaluating their proficiency in complex, open-ended tasks c…