语言大模型 / LLM - arXivDaily 专题

2505.04174 2025-05-21 cs.LG cs.AI cs.NI eess.SP 88%

On-Device LLM for Context-Aware Wi-Fi Roaming

Ju-Hyung Lee, Yanqing Lu, Klaus Doppler

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

2411.15982 2025-05-13 cs.AR cs.AI cs.LG 88%

Anda: Unlocking Efficient LLM Inference with a Variable-Length Grouped Activation Data Format

Chao Fang, Man Shi, Robin Geens, Arne Symons, Zhongfeng Wang, Marian Verhelst

机构 * School of Electronic Science and Engineering, Nanjing University, China（南京大学电子科学与技术学院）； ESAT-MICAS, KU Leuven, Belgium（KU莱顿大学ESAT-MICAS）

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；post-training(abstract)

Comments To appear in 2025 IEEE International Symposium on High-Performance Computer Architecture (HPCA 2025)

2409.09013 2025-04-29 cs.AI cs.CL 88%

AI-LieDar: Examine the Trade-off Between Utility and Truthfulness in LLM Agents

Zhe Su, Xuhui Zhou, Sanketh Rangreji, Anubha Kabra, Julia Mendelsohn, Faeze Brahman, Maarten Sap

机构 * Carnegie Mellon University（卡内基梅隆大学）； University of Michigan（密歇根大学）； Allen Institute for AI（人工智能阿伦研究所）

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

2504.14365 2025-04-22 cs.LG cs.AI cs.AR 88%

Accelerating LLM Inference with Flexible N:M Sparsity via A Fully Digital Compute-in-Memory Accelerator

Akshat Ramachandran, Souvik Kundu, Arnab Raha, Shamik Kundu, Deepak K. Mathaikutty, Tushar Krishna

机构 * Georgia Institute of Technology（佐治亚理工学院）； Intel Labs（英特尔实验室）； Intel Corporation（英特尔公司）

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

2410.10877 2025-03-07 cs.CL cs.AI 88%

Improving Data Efficiency via Curating LLM-Driven Rating Systems

Jinlong Pang, Jiaheng Wei, Ankit Parag Shah, Zhaowei Zhu, Yaxuan Wang, Chen Qian, Yang Liu, Yujia Bao, Wei Wei

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

2410.08115 2025-02-19 cs.CL cs.AI 88%

Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System

Weize Chen, Jiarui Yuan, Chen Qian, Cheng Yang, Zhiyuan Liu, Maosong Sun

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

Comments Under review

2502.11895 2025-02-18 cs.LG cs.AI 88%

Continual Quantization-Aware Pre-Training: When to transition from 16-bit to 1.58-bit pre-training for BitNet language models?

Jacob Nielsen, Peter Schneider-Kamp, Lukas Galke

专题命中效率与部署：language model(title,abstract)；LLM(abstract)；large language model(abstract)；post-training(abstract)

2502.04392 2025-02-10 cs.CL cs.AI 88%

Division-of-Thoughts: Harnessing Hybrid Language Model Synergy for Efficient On-Device Agents

Chenyang Shao, Xinyuan Hu, Yutang Lin, Fengli Xu

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

2501.15255 2025-01-28 cs.LG cs.AI 88%

Lightweight and Post-Training Structured Pruning for On-Device Large Lanaguage Models

Zihuai Xu, Yang Xu, Hongli Xu, Yunming Liao, Zhiwei Yao, Zuan Xie

专题命中效率与部署：post-training(title,abstract)；LLM(abstract)；large language model(abstract)；language model(abstract)

2412.10425 2025-01-13 cs.CL cs.AI 88%

Active Inference for Self-Organizing Multi-LLM Systems: A Bayesian Thermodynamic Approach to Adaptation

Rithvik Prakki

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

2412.03578 2024-12-06 cs.SE cs.AI cs.CL cs.PL 88%

PerfCodeGen: Improving Performance of LLM Generated Code with Execution Feedback

Yun Peng, Akhilesh Deepak Gotmare, Michael Lyu, Caiming Xiong, Silvio Savarese, Doyen Sahoo

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

2402.01093 2024-11-01 cs.LG cs.CL 88%

Need a Small Specialized Language Model? Plan Early!

David Grangier, Angelos Katharopoulos, Pierre Ablin, Awni Hannun

专题命中效率与部署：language model(title,abstract)；large language model(abstract)；small language model(abstract)；pretraining(abstract)

2409.15353 2024-09-25 eess.AS cs.CL cs.LG cs.SD 88%

Contextualization of ASR with LLM using phonetic retrieval-based augmentation

Zhihong Lei, Xingyu Na, Mingbin Xu, Ernest Pusateri, Christophe Van Gysel, Yuanyuan Zhang, Shiyi Han, Zhen Huang

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

2408.11609 2024-08-26 cs.CL cs.AI 88%

Xinyu: An Efficient LLM-based System for Commentary Generation

Yiquan Wu, Bo Tang, Chenyang Xi, Yu Yu, Pengyu Wang, Yifei Liu, Kun Kuang, Haiying Deng, Zhiyu Li, Feiyu Xiong, Jie Hu, Peng Cheng, Zhonghao Wang, Yi Wang, Yi Luo, Mingchuan Yang

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

2408.06610 2024-08-14 cs.CV cs.CL cs.LG 88%

CROME: Cross-Modal Adapters for Efficient Multimodal LLM

Sayna Ebrahimi, Sercan O. Arik, Tejas Nama, Tomas Pfister

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

2407.10582 2024-07-16 cs.CL cs.AI 88%

Boosting Zero-Shot Crosslingual Performance using LLM-Based Augmentations with Effective Data Selection

Barah Fazili, Ashish Sunil Agrawal, Preethi Jyothi

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

Comments Accepted in Findings of ACL 2024

2309.17382 2024-07-02 cs.AI cs.LG 88%

Reason for Future, Act for Now: A Principled Framework for Autonomous LLM Agents with Provable Sample Efficiency

Zhihan Liu, Hao Hu, Shenao Zhang, Hongyi Guo, Shuqi Ke, Boyi Liu, Zhaoran Wang

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

Comments Accepted by International Conference on Machine Learning (ICML) 2024

2405.17438 2024-05-29 cs.PL cs.AI cs.LG 88%

An LLM-Tool Compiler for Fused Parallel Function Calling

Simranjit Singh, Andreas Karatzas, Michael Fore, Iraklis Anagnostopoulos, Dimitrios Stamoulis

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

2306.12420 2024-05-07 cs.CL cs.AI 88%

LMFlow: An Extensible Toolkit for Finetuning and Inference of Large Foundation Models

Shizhe Diao, Rui Pan, Hanze Dong, Ka Shun Shum, Jipeng Zhang, Wei Xiong, Tong Zhang

专题命中效率与部署：foundation model(title,abstract)；language model(abstract)；instruction tuning(abstract)；pretraining(abstract)

Comments Published in NAACL 2024 Demo Track

2309.10105 2024-04-16 cs.CL cs.LG 88%

Understanding Catastrophic Forgetting in Language Models via Implicit Inference

Suhas Kotha, Jacob Mitchell Springer, Aditi Raghunathan

专题命中效率与部署：language model(title,abstract)；instruction tuning(abstract)；pretraining(abstract)；prompting(abstract)

Comments ICLR 2024

2402.14160 2024-03-06 cs.LG cs.AI 88%

Recursive Speculative Decoding: Accelerating LLM Inference via Sampling Without Replacement

Wonseok Jeon, Mukul Gagrani, Raghavv Goel, Junyoung Park, Mingu Lee, Christopher Lott

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

Comments 82 pages, 9 figures, 54 tables

2309.01885 2023-12-04 stat.ML cs.CL cs.LG 88%

QuantEase: Optimization-based Quantization for Language Models

Kayhan Behdin, Ayan Acharya, Aman Gupta, Qingquan Song, Siyu Zhu, Sathiya Keerthi, Rahul Mazumder

专题命中效率与部署：language model(title,abstract)；large language model(abstract)；post-training(abstract)；prompting(abstract)

2305.11186 2023-10-11 cs.CL cs.LG 88%

Compress, Then Prompt: Improving Accuracy-Efficiency Trade-off of LLM Inference with Transferable Prompt

Zhaozhuo Xu, Zirui Liu, Beidi Chen, Yuxin Tang, Jue Wang, Kaixiong Zhou, Xia Hu, Anshumali Shrivastava

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

2306.03078 2023-06-06 cs.CL cs.LG 88%

SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compression

Tim Dettmers, Ruslan Svirschevski, Vage Egiazarian, Denis Kuznedelev, Elias Frantar, Saleh Ashkboos, Alexander Borzunov, Torsten Hoefler, Dan Alistarh

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

Comments Extended preprint

2302.12173 2023-05-08 cs.CR cs.AI cs.CL cs.CY 88%

Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection

Kai Greshake, Sahar Abdelnabi, Shailesh Mishra, Christoph Endres, Thorsten Holz, Mario Fritz

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

2402.10517 2024-06-24 cs.LG 88%

Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs

Yeonhong Park, Jake Hyun, SangLyul Cho, Bonggeun Sim, Jae W. Lee

专题命中效率与部署：LLM(title,abstract)；large language model(abstract)；language model(abstract)；post-training(abstract)

Comments To appear at ICML 2024. Code is available at https://github.com/SNU-ARC/any-precision-llm

2401.05459 2024-05-09 cs.HC cs.AI cs.SE 88%

Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security

Yuanchun Li, Hao Wen, Weijun Wang, Xiangyu Li, Yizhen Yuan, Guohong Liu, Jiacheng Liu, Wenxing Xu, Xiang Wang, Yi Sun, Rui Kong, Yile Wang, Hanfei Geng, Jian Luan, Xuefeng Jin, Zilong Ye, Guanjing Xiong, Fan Zhang, Xiang Li, Mengwei Xu, Zhijun Li, Peng Li, Yang Liu, Ya-Qin Zhang, Yunxin Liu

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

Comments https://github.com/MobileLLM/Personal_LLM_Agents_Survey

2604.15794 2026-08-19 cs.LG cs.AI cs.CL 版本更新 88%

Self-Distillation as a Performance Recovery Mechanism for LLMs: Counteracting Compression and Catastrophic Forgetting

自蒸馏作为大语言模型的性能恢复机制：对抗压缩与灾难性遗忘

Chi Liu, Xin Chen, Xu Zhou, Fangbo Tu, Srinivasan Manoharan

机构 * PayPal AI

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

AI总结本文提出基于自蒸馏微调的性能恢复框架，通过理论分析和实验验证，证明自蒸馏能有效恢复模型能力，揭示了高维流形对齐与性能恢复之间的强相关性。

Comments 18 pages, 8 figures

2608.12253 2026-08-18 cs.CL cs.AI cs.LG 版本更新 88%

One Frozen Simulator Is Not Enough: Simulator Collapse in Multi-Agent RL

单个冻结模拟器不够：多智能体强化学习中的模拟器崩溃问题

Simon Yu, Nicholas Tomlin, Marwa Abdulhai, Ximing Lu, Derek Chong, Abe Hou, Dilara Soylu, Sergey Levine, Christopher D. Manning, Weiyan Shi

机构 * Northeastern University（东北大学）； New York University（纽约大学）； UC Berkeley（加州大学伯克利分校）； University of Washington（华盛顿大学）； Stanford University（斯坦福大学）

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

AI总结针对人机交互多智能体强化学习中单个LLM模拟器导致的策略泛化缺陷，提出Verbalized Sampling和Co-Training两种方案，在多轮基准测试和真实用户研究中显著提升了性能，发布了开源框架SCOPE。

Comments 42 pages, 29 figures

2504.06659 2026-08-18 cs.LG cs.AI cs.CL 版本更新 88%

Leveraging Machine Unlearning for Cost-Efficient Preference Alignment

利用机器遗忘实现高性价比的偏好对齐

Xiaohua Feng, Yuyuan Li, Huwei Ji, Jiaming Zhang, Li Zhang, Tianyu Du, Chaochao Chen

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

AI总结该研究针对现有偏好对齐方法成本高的问题，提出结合LLM遗忘的U2A框架，通过双层优化实现负示例的最优选择与遗忘，经实验验证有效。

Comments Accepted by ICML 2026. 12 pages, 6 figures, and 4 tables. Code available at https://github.com/muyiahhh/U2A

1. 效率与部署 22280 篇

On-Device LLM for Context-Aware Wi-Fi Roaming

Anda: Unlocking Efficient LLM Inference with a Variable-Length Grouped Activation Data Format

AI-LieDar: Examine the Trade-off Between Utility and Truthfulness in LLM Agents

Accelerating LLM Inference with Flexible N:M Sparsity via A Fully Digital Compute-in-Memory Accelerator

Improving Data Efficiency via Curating LLM-Driven Rating Systems

Optima: Optimizing Effectiveness and Efficiency for LLM-Based Multi-Agent System

Continual Quantization-Aware Pre-Training: When to transition from 16-bit to 1.58-bit pre-training for BitNet language models?

Division-of-Thoughts: Harnessing Hybrid Language Model Synergy for Efficient On-Device Agents

Lightweight and Post-Training Structured Pruning for On-Device Large Lanaguage Models

Active Inference for Self-Organizing Multi-LLM Systems: A Bayesian Thermodynamic Approach to Adaptation

PerfCodeGen: Improving Performance of LLM Generated Code with Execution Feedback

Need a Small Specialized Language Model? Plan Early!

Contextualization of ASR with LLM using phonetic retrieval-based augmentation

Xinyu: An Efficient LLM-based System for Commentary Generation

CROME: Cross-Modal Adapters for Efficient Multimodal LLM

Boosting Zero-Shot Crosslingual Performance using LLM-Based Augmentations with Effective Data Selection

Reason for Future, Act for Now: A Principled Framework for Autonomous LLM Agents with Provable Sample Efficiency

An LLM-Tool Compiler for Fused Parallel Function Calling

LMFlow: An Extensible Toolkit for Finetuning and Inference of Large Foundation Models

Understanding Catastrophic Forgetting in Language Models via Implicit Inference

Recursive Speculative Decoding: Accelerating LLM Inference via Sampling Without Replacement

QuantEase: Optimization-based Quantization for Language Models

Compress, Then Prompt: Improving Accuracy-Efficiency Trade-off of LLM Inference with Transferable Prompt

SpQR: A Sparse-Quantized Representation for Near-Lossless LLM Weight Compression

Not what you've signed up for: Compromising Real-World LLM-Integrated Applications with Indirect Prompt Injection

Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs

Personal LLM Agents: Insights and Survey about the Capability, Efficiency and Security

Self-Distillation as a Performance Recovery Mechanism for LLMs: Counteracting Compression and Catastrophic Forgetting

One Frozen Simulator Is Not Enough: Simulator Collapse in Multi-Agent RL

Leveraging Machine Unlearning for Cost-Efficient Preference Alignment