SLP-L2.2
SHRINKV: KEY-VALUE CACHE COMPRESSION WITH PROGRESSIVE HIDDEN STATES SHRINKING TO MITIGATE PREFILLING LATENCY
Jian Yuan, Shanghai Jiao Tong University, China; Ziwei He, Shanghai Innovation Institute, China; Zhouhan Lin, Bo Jiang, Shanghai Jiao Tong University, China
Session:
SLP-L2: Language Generation: Methods and Applications I Oral
Track:
Speech and Language Processing [SL]
Location:
Room 115
Presentation Time:
Tue, 5 May, 14:20 - 14:40
Presentation
Discussion
Resources
No resources available.
Session SLP-L2
SLP-L2.1: HIERARCHICAL ORTHOGONAL RESIDUAL SPREAD FOR PRECISE MASSIVE EDITING IN LARGE LANGUAGE MODELS
Xiaojie Gu, Independent Researcher, China; Guangxu Chen, UESTC, China; Yuheng Yang, Independent Researcher, China; Jingxin Han, Shanghai University, China; Andi Zhang, University of Manchester, China
SLP-L2.2: SHRINKV: KEY-VALUE CACHE COMPRESSION WITH PROGRESSIVE HIDDEN STATES SHRINKING TO MITIGATE PREFILLING LATENCY
Jian Yuan, Shanghai Jiao Tong University, China; Ziwei He, Shanghai Innovation Institute, China; Zhouhan Lin, Bo Jiang, Shanghai Jiao Tong University, China
SLP-L2.3: LET MORE EXPERTS SPEAK: BALANCING EXPLORATION AND EXPLOITATION IN PEFT FOR MIXTURE-OF-EXPERTS MODELS
Yi-Zeng Fang, Juinn-Dar Huang, National Yang Ming Chiao Tung University, Taiwan
SLP-L2.4: SATBADEDIT: TOWARDS EFFICIENT AND ROBUST MULTI-TRIGGER BACKDOOR INJECTION IN LARGE LANGUAGE MODELS
Yue Chen, Tianjin University, China; Zhao Xiaohu, Alibaba International Digital Commerce Group, China; Xinwei Wu, Jianxiang Peng, Dan Shi, Lei Yang, Tianjin University, China; Linlong Xu, Alibaba International Digital Commerce Group, China; Yueheng Sun, Deyi Xiong, Tianjin University, China
SLP-L2.5: PRESERVING KNOWLEDGE IN LARGE LANGUAGE MODELS VIA MODEL-AGNOSTIC INTRINSIC GENERATIVE REPLAY
Zilun Zhang, Yutao Sun, Zhejiang University, China; Tiancheng Zhao, Binjiang Research Institute of Zhejiang University, China; Leigang Sha, Zhejiang University, China; Ruochen Xu, Linker Technology Research Co. Ltd, China; Kyusong Lee, Binjiang Research Institute of Zhejiang University, China; Jianwei Yin, Zhejiang University, China
SLP-L2.6: PROACTIVE SAFETY DELIBERATION: GUIDING LARGE REASONING MODELS WITH DISTILLED PRINCIPLES
Yuxin zhou, Xiao Ding, Harbin Institute of Technology, China; Qi Shi, Tsinghua University, China; Ye He, Kai Xiong, Yijia Meng, Tianle Chang, Jinglong Gao, Harbin Institute of Technology, China
Contacts