语言大模型 / LLM - arXivDaily 专题

2405.04065 2025-06-16 cs.CL 85%

FlashBack:Efficient Retrieval-Augmented Language Modeling for Long Context Inference

Runheng Liu, Xingchen Xiao, Heyan Huang, Zewen Chi, Zhijing Wu

机构 * School of Computer Science and Technology, Beijing Institute of Technology（计算机科学与技术学院，北京理工大学）

专题命中效率与部署：language model(title,abstract)；LLM(abstract)；large language model(abstract)；分类 cs.CL

Comments ACL 2025 Findings, 14 pages

2506.09516 2025-06-12 stat.ML cs.LG stat.ME 85%

LLM-Powered CPI Prediction Inference with Online Text Time Series

Yingying Fan, Jinchi Lv, Ao Sun, Yurou Wang

机构 * Marshall School of Business, University of Southern California（马歇尔商学院，南加州大学）； Paula and Gregory Chow Institute for Studies in Economics, Xiamen University（陈保罗和格雷戈里·楚经济学研究所，厦门大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments 73 pages, 13 figures

2410.04790 2025-06-10 cs.CL 85%

PECAN: LLM-Guided Dynamic Progress Control with Attention-Guided Hierarchical Weighted Graph for Long-Document QA

Xinyu Wang, Yanzheng Xiang, Lin Gui, Yulan He

机构 * Department of Computer Science, University of Warwick（沃里克大学计算机科学系）； Department of Informatics, King’s College London（伦敦国王学院信息学系）； The Alan Turing Institute（艾伦·图灵研究所）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

2409.15104 2025-06-10 cs.DC cs.LG 85%

PecSched: Preemptive and Efficient Cluster Scheduling for LLM Inference

Zeyu Zhang, Haiying Shen

机构 * University of Virginia（弗吉尼亚大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

2506.04594 2025-06-06 cs.NI cs.AI eess.SP 85%

Intelligent Channel Allocation for IEEE 802.11be Multi-Link Operation: When MAB Meets LLM

Shumin Lian, Jingwen Tong, Jun Zhang, Liqun Fu

机构 * School of Informatics, Xiamen University（厦门大学信息学院）； Department of Electronic and Computer Engineering, The Hong Kong University of Science and Technology（香港科技大学电子与计算机工程系）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments This work has been accepted by JSAC 2025

2506.03700 2025-06-05 cs.CL 85%

AdaDecode: Accelerating LLM Decoding with Adaptive Layer Parallelism

Zhepei Wei, Wei-Lin Chen, Xinyu Zhu, Yu Meng

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments ICML 2025. Code: https://github.com/weizhepei/AdaDecode

2411.17089 2025-06-05 cs.LG cs.DC cs.PF 85%

KVPR: Efficient LLM Inference with I/O-Aware KV Cache Partial Recomputation

Chaoyi Jiang, Lei Gao, Hossein Entezari Zarch, Murali Annavaram

机构 * University of Southern California（南加州大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments ACL Findings 2025

2506.02490 2025-06-04 cs.DC cs.AI cs.SE 85%

Simplifying Root Cause Analysis in Kubernetes with StateGraph and LLM

Yong Xiang, Charley Peter Chen, Liyi Zeng, Wei Yin, Xin Liu, Hu Li, Wei Xu

机构 * Tsinghua University（清华大学）； Harmonic Inc（Harmonic公司）； Peng Cheng Laboratory（鹏城实验室）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments 12 pages, 13 figures, 5 tables

2506.01900 2025-06-03 cs.AI cs.CE cs.CR 85%

COALESCE: Economic and Security Dynamics of Skill-Based Task Outsourcing Among Team of Autonomous LLM Agents

Manish Bhatt, Ronald F. Del Rosario, Vineeth Sai Narajala, Idan Habler

机构 * Amazon（亚马逊）； SAP ； Amazon Web Services（亚马逊网络服务）； Intuit

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments 20 pages, 2 figures, github linked

2502.11471 2025-06-02 cs.CL cs.IR 85%

GLTW: Joint Improved Graph Transformer and LLM via Three-Word Language for Knowledge Graph Completion

Kangyang Luo, Yuzhuo Bai, Cheng Gao, Shuzheng Si, Yingli Shen, Zhu Liu, Zhitong Wang, Cunliang Kong, Wenhao Li, Yufei Huang, Ye Tian, Xuantang Xiong, Lei Han, Maosong Sun

机构 * Department of Computer Science and Technology, Tsinghua University（清华大学计算机科学与技术系）； Institute for AI, Tsinghua University（清华大学人工智能研究院）； Tencent Robotics X（腾讯机器人X）； Jiangsu Collaborative Innovation Center for Language Ability（江苏省语言能力协同创新中心）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments Accepted by ACL2025(Findings)

2505.23022 2025-05-30 cs.LG 85%

SCORPIO: Serving the Right Requests at the Right Time for Heterogeneous SLOs in LLM Inference

Yinghao Tang, Tingfeng Lan, Xiuqi Huang, Hui Lu, Wei Chen

机构 * State Key Lab of CAD&CG（计算机辅助设计与图形学国家重点实验室）； University of Virginia（弗吉尼亚大学）； The University of Texas at Arlington（德克萨斯大学阿灵顿分校）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

2505.22852 2025-05-30 cs.CR cs.AI 85%

Operationalizing CaMeL: Strengthening LLM Defenses for Enterprise Deployment

Krti Tallam, Emma Miller

机构 * SentinelAI

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

2505.22375 2025-05-30 cs.CL 85%

Pangu Embedded: An Efficient Dual-system LLM Reasoner with Metacognition

Hanting Chen, Yasheng Wang, Kai Han, Dong Li, Lin Li, Zhenni Bi, Jinpeng Li, Haoyu Wang, Fei Mi, Mingjian Zhu, Bin Wang, Kaikai Song, Yifei Fu, Xu He, Yu Luo, Chong Zhu, Quan He, Xueyu Wu, Wei He, Hailin Hu, Yehui Tang, Dacheng Tao, Xinghao Chen, Yunhe Wang

机构 * Huawei（华为）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

2505.16508 2025-05-30 cs.DC cs.AI cs.NI cs.PF 85%

Edge-First Language Model Inference: Models, Metrics, and Tradeoffs

SiYoung Jang, Roberto Morabito

机构 * Nokia Bell Labs Cambridge（诺基亚贝尔实验室剑桥）； EURECOM Biot（EURECOM比奥）

专题命中效率与部署：language model(title,abstract)；small language model(abstract)；SLM(abstract)；分类 cs.AI

Comments This paper has been accepted for publication and presentation at the 45th IEEE International Conference on Distributed Computing Systems (IEEE ICDCS 2025). The copyright will be transferred to IEEE upon publication in the conference proceedings

2505.21987 2025-05-29 cs.LG 85%

ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning

Zhendong Mi, Zhenglun Kong, Geng Yuan, Shaoyi Huang

机构 * Stevens Institute of Technology（史蒂文斯理工学院）； Harvard University（哈佛大学）； University of Georgia（佐治亚大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments 9 pages, 2 figures, 13 tables

2505.21919 2025-05-29 cs.ET cs.AI cs.DC 85%

Towards Efficient Key-Value Cache Management for Prefix Prefilling in LLM Inference

Yue Zhu, Hao Yu, Chen Wang, Zhuoran Liu, Eun Kyung Lee

机构 * IBM T. J. Watson Research Center（IBM T. J. Watson研究院）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments This paper has been accepted at IEEE Cloud 2025 as WIP paper. The final version will appear in IEEE Xplore

2504.12339 2025-05-29 cs.CL cs.SD eess.AS 85%

GOAT-TTS: Expressive and Realistic Speech Generation via A Dual-Branch LLM

Yaodong Song, Hongjie Chen, Jie Lian, Yuxin Zhang, Guangmin Xia, Zehan Li, Genliang Zhao, Jian Kang, Jie Li, Yongxiang Li, Xuelong Li

机构 * Institute of Artificial Intelligence (TeleAI), China Telecom（人工智能研究院（TeleAI），中国电信）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

2503.07320 2025-05-29 cs.HC cs.AI 85%

When Trust Collides: Decoding Human-LLM Cooperation Dynamics through the Prisoner's Dilemma

Guanxuan Jiang, Shirao Yang, Yuyang Wang, Pan Hui

机构 * The Hong Kong University of Science and Technology (Guangzhou)（香港科学与技术大学（广州））； The Hong Kong University of Science and Technology（香港科学与技术大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

2505.20438 2025-05-28 cs.CL 85%

HAMburger: Accelerating LLM Inference via Token Smashing

Jingyu Liu, Ce Zhang

机构 * Department of Computer Science University of Chicago（计算机科学系芝加哥大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

2505.19003 2025-05-27 cs.AI 85%

Aligning LLM with human travel choices: a persona-based embedding learning approach

Tianming Liu, Manzi Li, Yafeng Yin

机构 * Department of Civil and Environmental Engineering, University of Michigan, Ann Arbor, United States（密歇根大学土木与环境工程系）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments 32 pages, 8 figures

2502.11598 2025-05-27 cs.CL 85%

Can LLM Watermarks Robustly Prevent Unauthorized Knowledge Distillation?

Leyi Pan, Aiwei Liu, Shiyu Huang, Yijian Lu, Xuming Hu, Lijie Wen, Irwin King, Philip S. Yu

机构 * Tsinghua University（清华大学）； Zhipu AI（智谱AI）； The Hong Kong University of Science and Technology (Guangzhou)（香港科技大学（广州））； The Chinese University of Hong Kong（香港中文大学）； University of Illinois at Chicago（伊利诺伊大学香槟分校）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments Accepted by ACL 2025 (Main)

2502.11417 2025-05-27 cs.LG cs.DC 85%

DiSCo: Device-Server Collaborative LLM-Based Text Streaming Services

Ting Sun, Penghan Wang, Fan Lai

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校）； Purdue University（普渡大学）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments ACL 2025

2501.08192 2025-05-27 cs.AI cs.AR cs.DC 85%

PRESERVE: Prefetching Model Weights and KV-Cache in Distributed LLM Serving

Ahmet Caner Yüzügüler, Jiawei Zhuang, Lukas Cavigelli

机构 * Huawei Zurich Research Center Switzerland（华为苏黎世研究中心）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

2503.20641 2025-05-26 cs.CL 85%

Unlocking Efficient Long-to-Short LLM Reasoning with Model Merging

Han Wu, Yuxuan Yao, Shuqi Liu, Zehua Liu, Xiaojin Fu, Xiongwei Han, Xing Li, Hui-Ling Zhen, Tao Zhong, Mingxuan Yuan

专题命中效率与部署：LLM(title)；large language model(abstract)；language model(abstract)；SFT(abstract)

Comments Technical report

2505.16221 2025-05-23 cs.AI 85%

LightRouter: Towards Efficient LLM Collaboration with Minimal Overhead

Yifan Zhang, Xinkui Zhao, Zuxin Wang, Guanjie Cheng, Yueshen Xu, Shuiguang Deng, Jianwei Yin

机构 * School of Software Technology, Zhejiang University, Hangzhou, China（浙江大学软件技术学院）； School of Computer Science, Zhejiang University, Hangzhou, China（浙江大学计算机学院）； School of Software Engineering, Xidian University, Xi’an, China（西安电子科技大学软件工程学院）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.AI

2505.14417 2025-05-21 cs.CG cs.LG 85%

Towards Non-Euclidean Foundation Models: Advancing AI Beyond Euclidean Frameworks

Menglin Yang, Yifei Zhang, Jialin Chen, Melanie Weber, Rex Ying

机构 * Hong Kong University of Science and Technology (GZ) Guangzhou China（香港科技大学（广州））； Yale University New Haven United States（耶鲁大学）； Nanyang Technological University Singapore Singapore（南洋理工大学）； Harvard University Cambridge United States（哈佛大学）

专题命中效率与部署：foundation model(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments WWW 2025 Companion

2412.12504 2025-05-21 cs.IR cs.LG 85%

Boosting LLM-based Relevance Modeling with Distribution-Aware Robust Learning

Hong Liu, Saisai Gong, Yixin Ji, Kaixin Wu, Jia Xu, Jinjie Gu

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

Comments 8 pages

Journal ref CIKM(2024) 4718-4725

2404.17662 2025-05-21 cs.CL 85%

PLAYER*: Enhancing LLM-based Multi-Agent Communication and Interaction in Murder Mystery Games

Qinglin Zhu, Runcong Zhao, Bin Liang, Jinhua Du, Lin Gui, Yulan He

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

2505.13326 2025-05-20 cs.LG 85%

Thinking Short and Right Over Thinking Long: Serving LLM Reasoning Efficiently and Accurately

Yuhang Wang, Youhe Jiang, Bin Cui, Fangcheng Fu

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.LG

2505.12831 2025-05-20 cs.CL 85%

Contrastive Prompting Enhances Sentence Embeddings in LLMs through Inference-Time Steering

Zifeng Cheng, Zhonghui Wang, Yuchen Fu, Zhiwei Jiang, Yafeng Yin, Cong Wang, Qing Gu

机构 * State Key Laboratory for Novel Software Technology, Nanjing University, China（南京大学新型软件技术国家重点实验室）

专题命中效率与部署：prompting(title,abstract)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments ACL 2025

1. 效率与部署 22368 篇

FlashBack:Efficient Retrieval-Augmented Language Modeling for Long Context Inference

LLM-Powered CPI Prediction Inference with Online Text Time Series

PECAN: LLM-Guided Dynamic Progress Control with Attention-Guided Hierarchical Weighted Graph for Long-Document QA

PecSched: Preemptive and Efficient Cluster Scheduling for LLM Inference

Intelligent Channel Allocation for IEEE 802.11be Multi-Link Operation: When MAB Meets LLM

AdaDecode: Accelerating LLM Decoding with Adaptive Layer Parallelism

KVPR: Efficient LLM Inference with I/O-Aware KV Cache Partial Recomputation

Simplifying Root Cause Analysis in Kubernetes with StateGraph and LLM

COALESCE: Economic and Security Dynamics of Skill-Based Task Outsourcing Among Team of Autonomous LLM Agents

GLTW: Joint Improved Graph Transformer and LLM via Three-Word Language for Knowledge Graph Completion

SCORPIO: Serving the Right Requests at the Right Time for Heterogeneous SLOs in LLM Inference

Operationalizing CaMeL: Strengthening LLM Defenses for Enterprise Deployment

Pangu Embedded: An Efficient Dual-system LLM Reasoner with Metacognition

Edge-First Language Model Inference: Models, Metrics, and Tradeoffs

ACE: Exploring Activation Cosine Similarity and Variance for Accurate and Calibration-Efficient LLM Pruning

Towards Efficient Key-Value Cache Management for Prefix Prefilling in LLM Inference

GOAT-TTS: Expressive and Realistic Speech Generation via A Dual-Branch LLM

When Trust Collides: Decoding Human-LLM Cooperation Dynamics through the Prisoner's Dilemma

HAMburger: Accelerating LLM Inference via Token Smashing

Aligning LLM with human travel choices: a persona-based embedding learning approach

Can LLM Watermarks Robustly Prevent Unauthorized Knowledge Distillation?

DiSCo: Device-Server Collaborative LLM-Based Text Streaming Services

PRESERVE: Prefetching Model Weights and KV-Cache in Distributed LLM Serving

Unlocking Efficient Long-to-Short LLM Reasoning with Model Merging

LightRouter: Towards Efficient LLM Collaboration with Minimal Overhead

Towards Non-Euclidean Foundation Models: Advancing AI Beyond Euclidean Frameworks

Boosting LLM-based Relevance Modeling with Distribution-Aware Robust Learning

PLAYER*: Enhancing LLM-based Multi-Agent Communication and Interaction in Murder Mystery Games

Thinking Short and Right Over Thinking Long: Serving LLM Reasoning Efficiently and Accurately

Contrastive Prompting Enhances Sentence Embeddings in LLMs through Inference-Time Steering