Two tests are described that were developed for benchmarking and comparison of numerical codes in the context of AWAKE experiment.
生成讲解博客的中文版本
生成AI智能校验的中文版本