语言大模型 / LLM - arXivDaily 专题

2110.07577 2022-09-07 cs.CL cs.AI cs.LG 82%

UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning

Yuning Mao, Lambert Mathias, Rui Hou, Amjad Almahairi, Hao Ma, Jiawei Han, Wen-tau Yih, Madian Khabsa

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

Comments ACL 2022 (w. typo fixes)

2207.00112 2022-07-04 cs.LG cs.AI cs.CL 82%

Language model compression with weighted low-rank factorization

Yen-Chang Hsu, Ting Hua, Sungen Chang, Qian Lou, Yilin Shen, Hongxia Jin

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

Comments ICLR 2022

2205.08084 2022-05-20 cs.IR 82%

M6-Rec: Generative Pretrained Language Models are Open-Ended Recommender Systems

Zeyu Cui, Jianxin Ma, Chang Zhou, Jingren Zhou, Hongxia Yang

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

Comments 10 pages, 8 figures, proudly rejected by KDD 2022

2105.13880 2022-04-27 cs.CL cs.AI cs.LG 82%

Knowledge Inheritance for Pre-trained Language Models

Yujia Qin, Yankai Lin, Jing Yi, Jiajie Zhang, Xu Han, Zhengyan Zhang, Yusheng Su, Zhiyuan Liu, Peng Li, Maosong Sun, Jie Zhou

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

Comments NAACL 2022

2204.04813 2022-04-12 cs.CL cs.AI cs.LG 82%

Explanation Graph Generation via Pre-trained Language Models: An Empirical Study with Contrastive Learning

Swarnadeep Saha, Prateek Yadav, Mohit Bansal

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

Comments ACL 2022 (19 pages)

2112.01753 2022-03-22 cs.CL cs.AI cs.LG 82%

Probing Linguistic Information For Logical Inference In Pre-trained Language Models

Zeming Chen, Qiyue Gao

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

Comments AAAI 2022 camera ready version

2111.05754 2021-11-11 cs.CL cs.AI cs.LG 82%

Prune Once for All: Sparse Pre-Trained Language Models

Ofir Zafrir, Ariel Larey, Guy Boudoukh, Haihao Shen, Moshe Wasserblat

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

Comments ENLSP NeurIPS Workshop 2021, 12 pages

2110.08745 2021-10-19 cs.CL cs.AI cs.LG 82%

Reminding the Incremental Language Model via Data-Free Self-Distillation

Han Wang, Ruiliu Fu, Chengzhang Li, Xuejun Zhang, Jun Zhou, Yonghong Yan

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

Comments 8 pages, 5 figures

2110.08551 2021-10-19 cs.CL cs.AI cs.LG 82%

HRKD: Hierarchical Relational Knowledge Distillation for Cross-domain Language Model Compression

Chenhe Dong, Yaliang Li, Ying Shen, Minghui Qiu

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

Comments EMNLP 2021

2109.08359 2021-09-20 cs.CL cs.AI cs.LG 82%

Distilling Linguistic Context for Language Model Compression

Geondo Park, Gyeongman Kim, Eunho Yang

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

Comments EMNLP 2021. Code: https://github.com/GeondoPark/CKD

2104.08164 2021-09-10 cs.CL cs.AI cs.LG 82%

Editing Factual Knowledge in Language Models

Nicola De Cao, Wilker Aziz, Ivan Titov

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

Comments Accepted at EMNLP2021 Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing. Code at https://github.com/nicola-decao/KnowledgeEditor . 16 pages, 6 figures, 2 tables

2106.03613 2021-06-08 cs.CL cs.AI cs.LG 82%

RoSearch: Search for Robust Student Architectures When Distilling Pre-trained Language Models

Xin Guo, Jianlei Yang, Haoyi Zhou, Xucheng Ye, Jianxin Li

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

Comments 10 pages, 9 figures

2106.02302 2021-06-07 eess.AS cs.AI cs.CL cs.LG cs.SD 82%

Minimum Word Error Rate Training with Language Model Fusion for End-to-End Speech Recognition

Zhong Meng, Yu Wu, Naoyuki Kanda, Liang Lu, Xie Chen, Guoli Ye, Eric Sun, Jinyu Li, Yifan Gong

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

Comments 5 pages, Interspeech 2021

Journal ref Interspeech 2021, Brno, Czech Republic

1909.01136 2021-04-08 cs.CL cs.AI cs.LG 82%

Pre-training A Neural Language Model Improves The Sample Efficiency of an Emergency Room Classification Model

Binbin Xu, Cédric Gil-Jardiné, Frantz Thiessard, Eric Tellier, Marta Avalos, Emmanuel Lagarde

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

Comments Version of the published manuscript

Journal ref The 33rd Florida Artificial Intelligence Research Society Conference, 2020

2007.02629 2020-11-03 cs.CL cs.AI cs.LG 82%

Learning Spoken Language Representations with Neural Lattice Language Modeling

Chao-Wei Huang, Yun-Nung Chen

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

Comments Published in ACL 2020 as a short paper

1904.09408 2019-10-18 cs.CL cs.AI cs.LG 82%

Language Models with Transformers

Chenguang Wang, Mu Li, Alexander J. Smola

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

Comments 12 pages, 7 tables, 4 figures

1904.09324 2019-09-05 cs.CL cs.AI cs.LG stat.ML 82%

Mask-Predict: Parallel Decoding of Conditional Masked Language Models

Marjan Ghazvininejad, Omer Levy, Yinhan Liu, Luke Zettlemoyer

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

Comments EMNLP 2019

1806.06950 2018-06-20 cs.CL cs.AI cs.LG stat.ML 82%

GroupReduce: Block-Wise Low-Rank Approximation for Neural Language Model Shrinking

Patrick H. Chen, Si Si, Yang Li, Ciprian Chelba, Cho-jui Hsieh

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

2601.21169 2026-07-14 cs.CL cs.AI 82%

Output-Space Search: Targeting LLM Generations in a Frozen Encoder-Defined Output Space

输出空间搜索：在冻结编码器定义的输出空间中定向LLM生成

Tobias Materzok

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

AI总结 OS-Search通过在冻结编码器定义的输出空间中进行端点搜索，提升LLM生成的多样性和代码有效性。

Journal ref Proceedings of the First Workshop on Structured Understanding, Retrieval, and Generation in the LLM Era (SURGeLLM 2026), pp. 70-92, 2026

2605.12335 2026-05-13 cs.IR cs.AI cs.LG 82%

EHR-RAGp: Retrieval-Augmented Prototype-Guided Foundation Model for Electronic Health Records

EHR-RAGp: 用于电子健康记录的检索增强型原型引导基础模型

Saeed Shurrab, Mariam Al-Omari, Dana El Samad, Farah E. Shamout

机构 * New York University（纽约大学）； New York University Abu Dhabi（纽约大学阿布扎赫德）

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

AI总结本文提出EHR-RAGp模型，通过动态整合患者历史信息提升临床预测性能，优于现有基础模型和Transformer基线。

Comments Retrieval Augmented EHR Foundation Model

2601.17768 2026-02-02 cs.LG cs.AI cs.DC 82%

LLM-42: Enabling Determinism in LLM Inference with Verified Speculation

LLM-42: 通过验证推测实现LLM推理的确定性

Raja Gond, Aditya K Kamath, Ramachandran Ramjee, Ashish Panwar

机构 * Microsoft Research（微软研究院）； University of Washington（华盛顿大学）

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

AI总结 LLM-42通过验证推测方法，在LLM推理中实现确定性，利用形状一致的减少和轻量级验证-回滚循环，减少确定性需求的开销。

Comments https://github.com/microsoft/llm-42

2503.04971 2026-01-15 cs.LG cs.AI cs.DC cs.MA 82%

Incentivizing Multi-Tenant Split Federated Learning for Foundation Models at the Network Edge

在网络边缘激励多租户分割联邦学习以实现基础模型

Songyuan Li, Jia Hu, Geyong Min, Haojun Huang

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

AI总结本文提出PRINCE机制，通过价格激励协调多租户SFL租户，提升基础模型微调效率。

Comments Accepted for publication in IEEE/ACM Transactions on Networking. Index Terms: Foundation models, Edge computing, Split federated learning, Multi-tenant system, Incentive mechanism

2601.03482 2026-01-08 cs.AI cs.LG stat.AP 82%

Personalization of Large Foundation Models for Health Interventions

为健康干预个性化设计大型基础模型

Stefan Konigorski, Johannes E. Vedder, Babajide Alamu Owoyele, İbrahim Özkan

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

AI总结本文提出混合框架，结合LFMs和N-of-1试验，以实现个性化医疗中的因果推断与责任AI整合。

Comments Accepted to the AAAI 2026 Workshop on Personalization in the Era of Large Foundation Models (PerFM)

2510.00071 2025-10-13 cs.AI cs.CL 82%

ARS: Adaptive Reasoning Suppression for Efficient Large Reasoning Language Models

Dongqi Zheng

机构 * Independent Researcher（独立研究者）

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

Comments Accepted by 39th NeurIPS - Foundations of Reasoning in Language Models

2504.12316 2025-04-18 cs.CL cs.AI cs.CV 82%

Data Metabolism: An Efficient Data Design Schema For Vision Language Model

Jingyuan Zhang, Hongzhi Zhang, Zhou Haonan, Chenxi Sun, Xingguang ji, Jiakang Wang, Fanheng Kong, Yahui Liu, Qi Wang, Fuzheng Zhang

机构 * Kuaishou Technology（快手科技）

专题命中效率与部署：language model(title,abstract)；分类 cs.CL、cs.AI；foundation model(comments)

Comments To be presented at ICLR 2025, First Workshop on Open Science for Foundation Models

2408.15305 2024-08-29 cs.LG cs.AI cs.CV 82%

Parameter-Efficient Quantized Mixture-of-Experts Meets Vision-Language Instruction Tuning for Semiconductor Electron Micrograph Analysis

Sakhinana Sagar Srinivas, Chidaksh Ravuru, Geethan Sannidhi, Venkataramana Runkana

专题命中效率与部署：instruction tuning(title)；language model(abstract)；分类 cs.AI、cs.LG；foundation model(comments)

Comments Paper published at ICML 2024 Workshop on Foundation Models in the Wild

2402.05964 2024-04-09 cs.LG cs.CL cs.CV 82%

A Survey on Transformer Compression

Yehui Tang, Yunhe Wang, Jianyuan Guo, Zhijun Tu, Kai Han, Hailin Hu, Dacheng Tao

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

Comments Model Compression, Transformer, Large Language Model, Large Vision Model, LLM

1708.05963 2019-04-09 stat.ML cs.CL cs.LG cs.NE 82%

Neural Networks Compression for Language Modeling

Artem M. Grachev, Dmitry I. Ignatov, Andrey V. Savchenko

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

Comments Keywords: LSTM, RNN, language modeling, low-rank factorization, pruning, quantization. Published by Springer in the LNCS series, 7th International Conference on Pattern Recognition and Machine Intelligence, 2017

2608.27439 2026-08-28 cs.CR cs.AI 新提交 81%

RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution

RedEvoAgent：具备经验驱动技能演化的自动红队代理

Junjie Zhang, Hui Liu, Kecheng Chen, Xianbo Mo, Changsheng Chen, Haoliang Li

机构 * City University of Hong Kong（香港城市大学）； Shenzhen MSU-BIT University（深圳北理莫斯科大学）

专题命中效率与部署：LLM(summary_cn,abstract)；分类 cs.AI

AI总结针对LLM代理的越狱攻击风险，提出RedEvoAgent黑盒红队代理，通过提炼攻击轨迹实现技能演化，在多基准实验中性能优于基线，工具效率更高且可迁移。

2608.27244 2026-08-28 cs.DB cs.AI 新提交 81%

Compositional Online Learning for Semantic Data Processing Systems

面向语义数据处理系统的组合式在线学习

Paweł Liskowski, Fuheng Zhao, Benjamin Han, Anupam Datta, Dimitris Tsirogiannis

机构 * Snowflake Inc.（斯诺flake公司）

专题命中效率与部署：LLM(summary_cn,abstract)；分类 cs.AI

AI总结针对语义数据处理系统中LLM调用成本高、延迟大的问题，提出组合式在线学习框架，在Cortex AISQL案例中实现约8倍的实际性能提升。

1. 效率与部署 22497 篇

UniPELT: A Unified Framework for Parameter-Efficient Language Model Tuning

Language model compression with weighted low-rank factorization

M6-Rec: Generative Pretrained Language Models are Open-Ended Recommender Systems

Knowledge Inheritance for Pre-trained Language Models

Explanation Graph Generation via Pre-trained Language Models: An Empirical Study with Contrastive Learning

Probing Linguistic Information For Logical Inference In Pre-trained Language Models

Prune Once for All: Sparse Pre-Trained Language Models

Reminding the Incremental Language Model via Data-Free Self-Distillation

HRKD: Hierarchical Relational Knowledge Distillation for Cross-domain Language Model Compression

Distilling Linguistic Context for Language Model Compression

Editing Factual Knowledge in Language Models

RoSearch: Search for Robust Student Architectures When Distilling Pre-trained Language Models

Minimum Word Error Rate Training with Language Model Fusion for End-to-End Speech Recognition

Pre-training A Neural Language Model Improves The Sample Efficiency of an Emergency Room Classification Model

Learning Spoken Language Representations with Neural Lattice Language Modeling

Language Models with Transformers

Mask-Predict: Parallel Decoding of Conditional Masked Language Models

GroupReduce: Block-Wise Low-Rank Approximation for Neural Language Model Shrinking

Output-Space Search: Targeting LLM Generations in a Frozen Encoder-Defined Output Space

EHR-RAGp: Retrieval-Augmented Prototype-Guided Foundation Model for Electronic Health Records

LLM-42: Enabling Determinism in LLM Inference with Verified Speculation

Incentivizing Multi-Tenant Split Federated Learning for Foundation Models at the Network Edge

Personalization of Large Foundation Models for Health Interventions

ARS: Adaptive Reasoning Suppression for Efficient Large Reasoning Language Models

Data Metabolism: An Efficient Data Design Schema For Vision Language Model

Parameter-Efficient Quantized Mixture-of-Experts Meets Vision-Language Instruction Tuning for Semiconductor Electron Micrograph Analysis

A Survey on Transformer Compression

Neural Networks Compression for Language Modeling

RedEvoAgent: Automatic Red-Teaming Agent with Experience-Driven Skill Evolution

Compositional Online Learning for Semantic Data Processing Systems