Xie Yu

Xie Yu

Auto-generated profile, based on publicly available information

Xie Yu

School of Computer Science and Engineering, Beihang University, Beijing, China and MOE Engineering Research Center of Advanced Computer Application Technology, Beihang University, Beijing, China

Research Focus & Metrics

Academic Career Length

2Years (2025-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
22 April 2025
Verified
Conference Paper
Open access
YFYu Feng
Yu Feng
YGYangli-ao Geng
Yangli-ao Geng
Yifan ZhuYifan Zhu
Yifan Zhu
ZHZongfu Han
Zongfu Han
XYXie Yu
Xie Yu
MSMeina Song
Meina Song

8 August 2026
Verified
Conference Paper
Open access
DPDayan Pan
Dayan Pan
HSHongkang Su
Hongkang Su
JWJingyuan Wang
Jingyuan Wang
XYXie Yu
Xie Yu

8 August 2026
Verified
Conference Paper
Open access
DPDayan Pan
Dayan Pan
JWJingyuan Wang
Jingyuan Wang
XYXie Yu
Xie Yu
Parameter-efficient fine-tuning (PEFT) has become a standard approach for adapting large language models to downstream tasks. However, most existing PEFT methods rely on uniform and static adaptations, without accounting for the structured heterogeneity of attention across dimensions, heads, layers, and input tokens. In practice, attention representations exhibit non-uniform behavior, and positional encoding mechanisms such as rotary positional embeddings (RoPE) induce dimension-dependent positional structure, making uniform adaptation suboptimal. In this work, we propose DyPAM (Dynamic Positional Attention Modulation), a PEFT method that adapts how positional information contributes to attention by operating directly on the query and key representations. DyPAM combines input-conditioned, dimension-wise modulation with head-wise and layer-wise structural modulation, performing fine-grained adaptation of positional attention aligned with the RoPE-induced structure without modifying the pretrained backbone. Extensive experiments on mathematical and commonsense reasoning benchmarks across multiple backbone models demonstrate that DyPAM consistently outperforms existing strong PEFT baselines. The code is available for reproducibility at https://github.com/Beihang-BIGSCity/DyPAM.