PEAR: Position-Embedding-Agnostic Attention Re-weighting Enhances Retrieval-Augmented Generation with Zero Inference Overhead.

Tao Tan|Yining Qian|Ang Lv|Hongzhan Lin|Songhao Wu|Yongbo Wang|Feng Wang|Jingtong Wu|Xin Lu|Rui Yan


Anthology ID:DBLP:conf/www/TanQL0WWWWL025
Volume:Proceedings of the ACM on Web Conference 2025, WWW 2025, Sydney, NSW, Australia, 28 April 2025- 2 May 2025
Year:2025
Venue:The Web Conference (WWW)
Publisher:ACM
Pages:1693-1702
URL:https://doi.org/10.1145/3696410.3714795
DOI:https://doi.org/10.1145/3696410.3714795
DBLP:conf/www/TanQL0WWWWL025
BibTeX:
@inproceedings{tan-2025-pear, author = {Tao Tan and Yining Qian and Ang Lv and Hongzhan Lin and Songhao Wu and Yongbo Wang and Feng Wang and Jingtong Wu and Xin Lu and Rui Yan}, editor = {Guodong Long and Michale Blumestein and Yi Chang and Liane Lewin-Eytan and Zi Huang and Elad Yom-Tov}, title = {{PEAR: Position-Embedding-Agnostic Attention Re-weighting Enhances Retrieval-Augmented Generation with Zero Inference Overhead}}, booktitle = {{Proceedings of the ACM on Web Conference 2025, WWW 2025, Sydney, NSW, Australia, 28 April 2025- 2 May 2025}}, pages = {1693--1702}, publisher = {ACM}, year = {2025}, url = {https://doi.org/10.1145/3696410.3714795}, doi = {https://doi.org/10.1145/3696410.3714795} }