Dialectics of Alignment: Harnessing Unsafe Knowledge for Dynamic Safety Routing
对齐的辩证法:利用不安全知识实现动态安全路由
Maryam Hashemzadeh, Jerry Huang, Minseon Kim, Marc-Alexandre Côté, Sarath Chandar
机构
*
Chandar Research Lab(Chandar研究实验室)
;
Mila – Quebec AI Institute(魁北克AI研究所)
;
Université de Montréal(蒙特利尔大学)
;
Microsoft Research(微软研究院)
;
Polytechnique Montréal(蒙特利尔理工学院)
;
Canada CIFAR AI Chair(加拿大CIFAR人工智能主席)
专题命中
指令微调
:LLM(abstract,abstract_cn);large language model(abstract);language model(abstract);分类 cs.LG
CommentsThis paper has been significantly revised, expanded, and superseded by a more comprehensive version available at arXiv:2604.18738. The authors have chosen to withdraw this version to avoid overlap and direct readers to the updated work
UniNote: A Unified Embedding Model for Multimodal Representation and Ranking
UniNote: 一种用于多模态表示和排序的统一嵌入模型
Jinghan Zhao, Wenwei Jin, Anqi Li, Jintao Tong, Luya Mo, Jiawei Li, Bin Li, Yao Hu
机构
*
Xiaohongshu Beijing China(小红书北京中国)
;
Shanghai Jiao Tong University(上海交通大学)
;
Huazhong University of Science and Technology(华中科技大学)
;
Beijing Institute of Technology(北京理工大学)
BERT4beam: Large AI Model Enabled Generalized Beamforming Optimization
BERT4beam: 大型AI模型实现通用波束赋形优化
Yuhang Li, Yang Lu, Wei Chen, Bo Ai, Zhiguo Ding
机构
*
State Key Laboratory of Advanced Rail Autonomous Operation(先进轨道交通自主运行国家重点实验室)
;
School of Computer Science and Technology(计算机科学与技术学院)
;
School of Electronics and Information Engineering(电子与信息工程学院)
;
School of Electrical and Electronic Engineering (EEE)(电子与电气工程学院)
专题命中
指令微调
:LLM(abstract_cn);large language model(abstract);language model(abstract);分类 cs.LG
Comments20 pages Research paper on Packet Language Models for Network Intrusion Detection Systems(Without Deep Packet Inspection).Code available on GitHub
GPTQ-intrinsic LoRA: A Near-optimal Algorithm for Low-precision Quantization with Low-rank Adaptation
GPTQ-intrinsic LoRA: 一种用于低秩自适应低精度量化的近最优算法
Shihao Zhang, Rayan Saab
机构
*
Department of Mathematics, University of California San Diego(数学系,加州大学圣地亚哥分校)
;
Department of Mathematics and Halıcıoğlu Data Science Institute, University of California San Diego(数学系和Halıcıoğlu数据科学研究所,加州大学圣地亚哥分校)
Not All Explanations Simulate Equally: Comparing Verbalized Feature Attributions and Self-Generated Rationales
并非所有解释都能同等模拟:比较言语化特征归因与自生成理由
Pingjun Hong, Benjamin Roth
机构
*
Faculty of Computer Science, University of Vienna(维也纳大学计算机科学系)
;
UniVie Doctoral School Computer Science, University of Vienna(维也纳大学计算机科学博士学院)
;
Faculty of Philological and Cultural Studies, University of Vienna(维也纳大学文学与文化研究系)