Disentangle-then-Refine: LLM-Guided Decoupling and Structure-Aware Refinement for Graph Contrastive Learning
解耦后再细化:基于LLM的解耦与结构感知细化用于图对比学习
Zhaoxing Li, Hai-Feng Zhang, Xiaoming Zhang
机构
*
Institute of Physical Science and Information Technology(物理科学与信息技术研究院)
;
School of Mathematical Sciences(数学科学学院)
;
Qinghai Institute of Science and Technology Information(青海科学技术信息研究院)
专题命中
效率与部署
:LLM(title,title_cn);large language model(abstract);language model(abstract);分类 cs.AI
Route to Rome Attack: Directing LLM Routers to Expensive Models via Adversarial Suffix Optimization
通往罗马的攻击:通过对抗性后缀优化引导LLM路由器选择昂贵模型
Haochun Tang, Yuliang Yan, Jiahua Lu, Huaxiao Liu, Enyan Dai
机构
*
Key Laboratory of Symbolic Computation and Knowledge Engineering, MoE, Jilin University(符号计算与知识工程重点实验室,MoE,吉林大学)
;
The Hong Kong University of Science and Technology (Guangzhou)(香港科技大学(广州))
Zihong Zhang, Zuchao Li, Lefei Zhang, Ping Wang, Hai Zhao
机构
*
School of Artificial Intelligence, Wuhan University(武汉大学人工智能学院)
;
School of Computer Science, Wuhan University(武汉大学计算机学院)
;
School of Information Management, Wuhan University(武汉大学信息管理学院)
;
School of Computer Science, Shanghai Jiao Tong University(上海交通大学计算机学院)
专题命中
效率与部署
:LLM(abstract,abstract_cn);large language model(abstract);language model(abstract);分类 cs.CL、cs.AI
Commentsv2. New methodology, instead of semi supervised now focused on linear separability. Methodological updates, performance upgrades. The concept is still the same, but the experimental setup and the implementation details have changed