1 paper
Yangyifan Xu, Jinliang Lu, Jiajun Zhang
Ensembling different large language models (LLMs) to unleash their complementary potential and harness their individual strengths is highly valuable. Nevertheless, vocabulary discr…