作者位次 #9/10ICML 2026Poster· Deep Learning / Large Language Models
Training-Trajectory-Aware Token Selection
Zhanming Shen · Jiaqi Hu · Zeyu Qin · Hao Chen · Wentao Ye · Zenan Huang · Yihong Zhuang · Guoshan Lu · Junlin Zhou · Junbo Zhao
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。
Zhanming Shen · Jiaqi Hu · Zeyu Qin · Hao Chen · Wentao Ye · Zenan Huang · Yihong Zhuang · Guoshan Lu · Junlin Zhou · Junbo Zhao
Yangyang Zhong · Yanmei Gu · Zhengqing Zang · Xiaomeng Li · Yuqi Ding · Xibei Jia · Yuting Shen · Zhenzhong Lan · Liwang Zhu · Weiping Liu · Junlin Zhou · Haisheng Liu · Zhong Xin Yu · Pengxin Luo · Donglian Qi · Yunfeng Yan · Junbo Zhao
Masked Diffusion Language Models (MDLMs) promise parallel token generation and arbitrary-order decoding, yet it remains unclear to what extent current models truly realize these capabilities. We characterize MDLM behavior along two dimensions—parallelism strength and generation order—using Averag...