作者位次 #2/4ICML 2026Poster· Deep Learning / Generative Models and Autoencoders
Concept-Guided Tokenization: Closing the Gap Between Reconstruction and Generation
Yunqiao Yang · Haokun Lin · Guanzhong Wu · Ying Wei
数据与预训练OpenReview
首次获取该候选人的完整档案及 StartLux 评估共 20 积分；每位用户单独解锁，之后查看免费。主动刷新 20 积分。
实时查找电子邮箱，费用不超过20积分；未找到邮箱不扣费。
Yunqiao Yang · Haokun Lin · Guanzhong Wu · Ying Wei
Shujun Xia · Haokun Lin · Yichen WU · Yinan Zhou · Zixuan Li · Zhongwei Wan · Xingrun Xing · Yefeng Zheng · Xiang Li · Caifeng Shan · Zhenan Sun · Quanzheng Li
Jingxuan Zhang · Yunta Hsieh · Zhongwei Wan · Haokun Lin · Xin Wang · Ziqi Wang · Yingtie Lei · Mi Zhang
Xingrun Xing · Zheng Liu · Shitao Xiao · Boyan Gao · Yiming Liang · Haokun Lin · Xianlin Zeng · Guoqi Li · Jiajun Zhang
Modern large language models (LLMs) driven by scaling laws achieve emergent intelligence in large model sizes. Recently, the increasing concerns about cloud costs, latency and privacy make it an urgent requirement to develop compact edge language models. Distinguished from direct pretraining that...