Researcher — paper.dinq.me

作者位次 #10/16ICLR 2026Poster· Reinforcement Learning / Deep RL

RECAST: Expanding the Boundaries of LLMs' Complex Instruction Following with Multi-Constraint Data

Zhengkang Guo · Wenhao Liu · Mingchen Xie · Jingwen Xu · Zisu Huang · Muzhao Tian · Jianhan Xu · Yuanzhe Shen · Qi Qian · Muling Wu · Xiaohua Wang · Heda Wang · Yao Hu · Changze Lv · Xuanjing Huang · Xiaoqing Zheng

强化学习与后训练OpenReview

作者位次 #7/12ACL 2026Findings· Findings of ACL

Mitigating Position Bias in Transformers via Layer-Specific Positional Embedding Scaling

Zhenghua Wang · Yiran Ding · Changze Lv · Yixin Wu · Tianlong Li · Zhibo Xu · Muling Wu · Tianyuan Shi · Shizheng Li · Qi Qian · Xuanjing Huang · Xiaoqing Zheng

摘要

Large Language Models (LLMs) still struggle with the “lost-in-the-middle” problem, where critical information located in the middle of long-context inputs is often underrepresented or lost. While existing methods attempt to address this by combining multi-scale rotary position embeddings (RoPE),...

效率与系统检索与知识长上下文与架构OpenReview 项目页

Muling Wu

获取完整研究档案

联系方式

2026 顶会论文

RECAST: Expanding the Boundaries of LLMs' Complex Instruction Following with Multi-Constraint Data

Mitigating Position Bias in Transformers via Layer-Specific Positional Embedding Scaling

常见合作者