作者位次 #3/8ICML 2026Poster· Deep Learning / Large Language Models
MergeMix: Optimizing Mid-Training Data Mixtures via Learnable Model Merging
Jiapeng Wang · Changxin Tian · Kunlong Chen · ziqi liu · Jiaxin Mao · Xin Zhao · Zhiqiang Zhang · JUN ZHOU
数据与预训练OpenReview