MMLU-Pro: A More Robust and Challenging Multi-Task Language Understanding Benchmark

Wang, Yubo, Ma, Xueguang, Zhang, Ge · arXiv · 2024