语言大模型 / LLM - arXivDaily 专题

2108.10561 2021-08-25 cs.CL cs.AI cs.LG 67%

Taming the Beast: Learning to Control Neural Conversational Models

Andrea Madotto

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments PhD thesis

2104.01563 2021-04-06 cs.CL cs.AI cs.LG 67%

ReCAM@IITK at SemEval-2021 Task 4: BERT and ALBERT based Ensemble for Abstract Word Prediction

Abhishek Mittal, Ashutosh Modi

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Accepted at SemEval 2021 Task 4, 8 Pages (7 Pages main content + 1 pages for references)

2103.05552 2021-03-10 cs.CL cs.AI cs.LG 67%

Comparing Approaches to Dravidian Language Identification

Tommi Jauhiainen, Tharindu Ranasinghe, Marcos Zampieri

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Accepted to VarDial 2021 @ EACL 2021

2101.08523 2021-01-22 cs.CL cs.AI cs.LG 67%

Adv-OLM: Generating Textual Adversaries via OLM

Vijit Malik, Ashwani Bhat, Ashutosh Modi

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 5 Pages + 1 Page references + 3 Pages Appendix, Accepted at EACL 2021

2101.01761 2021-01-07 cs.LG cs.AI cs.CL cs.CV 67%

AutoDropout: Learning Dropout Patterns to Regularize Deep Networks

Hieu Pham, Quoc V. Le

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Accepted to AAAI 2021

2012.15416 2021-01-01 cs.CL cs.AI cs.LG 67%

Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation

Damian Pascual, Beni Egressy, Florian Bolli, Roger Wattenhofer

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Preprint. Work in progress

2012.05684 2020-12-11 cs.LG cs.AI cs.CL 67%

Recurrent Point Review Models

Kostadin Cvejoski, Ramses J. Sanchez, Bogdan Georgiev, Christian Bauckhage, Cesar Ojeda

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 8 pages, 6 figures, Published in: 2020 International Joint Conference on Neural Networks (IJCNN)

Journal ref 2020 International Joint Conference on Neural Networks (IJCNN), Glasgow, United Kingdom, 2020, pp. 1-8

2010.02986 2020-11-22 cs.CL cs.AI cs.LG 67%

Compositional Demographic Word Embeddings

Charles Welch, Jonathan K. Kummerfeld, Verónica Pérez-Rosas, Rada Mihalcea

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments To appear at EMNLP 2020

2010.06775 2020-10-15 cs.CL cs.AI cs.CV cs.LG 67%

Vokenization: Improving Language Understanding with Contextualized, Visual-Grounded Supervision

Hao Tan, Mohit Bansal

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments EMNLP 2020 (15 pages)

2010.01794 2020-10-13 cs.CL cs.AI cs.LG 67%

GenAug: Data Augmentation for Finetuning Text Generators

Steven Y. Feng, Varun Gangal, Dongyeop Kang, Teruko Mitamura, Eduard Hovy

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments EMNLP 2020 Deep Learning Inside Out (DeeLIO) Workshop; Code available at https://github.com/styfeng/GenAug

2009.07243 2020-09-16 cs.CL cs.AI cs.LG 67%

A Systematic Characterization of Sampling Algorithms for Open-ended Language Generation

Moin Nadeem, Tianxing He, Kyunghyun Cho, James Glass

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments To appear at AACL 2020; 9 pages, 12 figures, 2 tables

2005.07064 2020-05-15 cs.CL cs.AI cs.LG 67%

Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learning

Angeliki Lazaridou, Anna Potapenko, Olivier Tieleman

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments to appear at ACL 2020

2005.00813 2020-05-05 cs.CL cs.AI cs.LG 67%

Social Biases in NLP Models as Barriers for Persons with Disabilities

Ben Hutchinson, Vinodkumar Prabhakaran, Emily Denton, Kellie Webster, Yu Zhong, Stephen Denuyl

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments ACL 2020 short paper. 5 pages

Journal ref ACL 2020

2004.05569 2020-04-14 cs.CL cs.AI cs.LG 67%

Explaining Question Answering Models through Text Generation

Veronica Latcinnik, Jonathan Berant

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

1903.10145 2019-06-12 cs.LG cs.AI cs.CL cs.CV stat.ML 67%

Cyclical Annealing Schedule: A Simple Approach to Mitigating KL Vanishing

Hao Fu, Chunyuan Li, Xiaodong Liu, Jianfeng Gao, Asli Celikyilmaz, Lawrence Carin

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Published in NAACL 2019; The first two authors contribute equally; Code: https://github.com/haofuml/cyclical_annealing

1906.04043 2019-06-11 cs.CL cs.AI cs.HC cs.LG 67%

GLTR: Statistical Detection and Visualization of Generated Text

Sebastian Gehrmann, Hendrik Strobelt, Alexander M. Rush

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments ACL 2019 Demo Track

1812.06705 2018-12-18 cs.CL cs.AI cs.LG 67%

Conditional BERT Contextual Augmentation

Xing Wu, Shangwen Lv, Liangjun Zang, Jizhong Han, Songlin Hu

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments 9 pages, 1 figure

1612.09161 2017-08-08 cs.CV 67%

Learning Visual N-Grams from Web Data

Ang Li, Allan Jabri, Armand Joulin, Laurens van der Maaten

专题命中其他LLM ：language model(abstract)；prompting(abstract)

1611.01576 2016-11-22 cs.NE cs.AI cs.CL cs.LG 67%

Quasi-Recurrent Neural Networks

James Bradbury, Stephen Merity, Caiming Xiong, Richard Socher

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

Comments Submitted to conference track at ICLR 2017

1508.04395 2016-03-16 cs.CL cs.AI cs.LG cs.NE 67%

End-to-End Attention-based Large Vocabulary Speech Recognition

Dzmitry Bahdanau, Jan Chorowski, Dmitriy Serdyuk, Philemon Brakel, Yoshua Bengio

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.AI、cs.LG

2604.04820 2026-04-07 cs.AI cs.CL 66%

ANX: Protocol-First Design for AI Agent Interaction with a Supporting 3EX Decoupled Architecture

ANX：面向AI代理交互的协议优先设计及其支持的3EX解耦架构

Xu Mingze

机构 * Hangzhou Ziyou Data Technology Co., Ltd.（杭州自由数据科技有限公司）

专题命中其他LLM ：LLM(abstract,comments)；分类 cs.CL、cs.AI

AI总结本文提出ANX协议，通过协议创新、架构优化和工具补充，解决AI代理交互中高消耗、碎片化、安全性不足等问题，其核心创新包括代理原生设计、人机交互、轻量应用和可执行SOP。

Comments This open-source AI agent interaction protocol (ANX) is benchmarked against existing protocols (MCP, A2A, ANP, OpenCLI, SkillWeaver, CHEQ, COLLAB-LLM) across four dimensions: tooling, discovery, security, and multi-agent SOP collaboration. Code: https://github.com/mountorc/anx-protocol

2502.17304 2025-02-25 cs.CL cs.AI 66%

Child vs. machine language learning: Can the logical structure of human language unleash LLMs?

Uli Sauerland, Celia Matthaei, Felix Salfner

机构 * ZAS（莱布尼茨语言信息中心）； HU Berlin（柏林洪堡大学）

专题命中其他LLM ：LLM(abstract)；分类 cs.CL、cs.AI；language model(comments)

Comments ISCA/ITG Workshop on Diversity in Large Speech and Language Models

2404.15406 2024-05-24 cs.CV cs.AI cs.CL cs.MM 66%

Wiki-LLaVA: Hierarchical Retrieval-Augmented Generation for Multimodal LLMs

Davide Caffagni, Federico Cocchi, Nicholas Moratelli, Sara Sarto, Marcella Cornia, Lorenzo Baraldi, Rita Cucchiara

专题命中其他LLM ：LLM(abstract)；分类 cs.CL、cs.AI；foundation model(comments)

Comments CVPR 2024 Workshop on What is Next in Multimodal Foundation Models

2211.15363 2024-05-14 cs.CL cs.CR cs.DB cs.LG cs.SE 66%

On the Security Vulnerabilities of Text-to-SQL Models

Xutan Peng, Yipeng Zhang, Jingfeng Yang, Mark Stevenson

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.LG；LLM(comments)

Comments Best Paper Candidate at ISSRE 2023. Replaced "PLM" with "LLM" for better visibility

2509.02890 2025-09-30 cs.IR cs.AI 65%

Grocery to General Merchandise: A Cross-Pollination Recommender using LLMs and Real-Time Cart Context

Akshay Kekuda, Murali Mohana Krishna Dandu, Rimita Lahiri, Shiqin Cai, Sinduja Subramaniam, Evren Korpeoglu, Kannan Achan

机构 * Walmart Global Tech（沃尔玛全球技术）

专题命中其他LLM ：LLM(abstract)；分类 cs.AI；large language model(comments)；language model(comments)

Comments Accepted at RecSys 2025 EARL Workshop on Evaluating and Applying Recommender Systems with Large Language Models

2608.26973 2026-08-28 cs.CL cs.LG 新提交 62%

Squeezing More from Limited Data with Recursive Transformers

用递归Transformer从有限数据中挖掘更多价值

Serdar Gülbahar, Lukas Edman, Alexander Fraser

机构 * Technical University of Munich（慕尼黑工业大学）

专题命中其他LLM ：language model(abstract)；分类 cs.CL、cs.LG

AI总结该研究针对有限数据预训练场景，提出递归Transformer结合因式分解词嵌入的方法，在10M和100M词预算下性能优于标准Transformer，且与2025年BabyLM挑战赛优胜者表现相当。

2608.26506 2026-08-28 cs.LG cs.CL 新提交 62%

A Single Suffix to Break Them All: Basin-Aware Jailbreaks for Merged Model Families

一个后缀突破所有：针对合并模型家族的感知 Basin 越狱攻击

Yu Zhe, Yixin Tan, Junhao Wei, Wang Chen

机构 * RIKEN AIP（理化学研究所先进智能项目）； Institute of Science Tokyo（东京科学大学）； Zhejiang University（浙江大学）

专题命中其他LLM ：foundation model(abstract)；分类 cs.CL、cs.LG

AI总结该研究针对合并模型家族提出BAJ方法，利用预训练基础模型的越狱风险，通过最小-最大优化生成可迁移的对抗性后缀，在多种设置下均实现高迁移成功率且能抵御现有防御。

Comments Accepted by EMNLP findings 2026

2608.23623 2026-08-26 cs.SE cs.AI cs.LG 新提交 62%

When May an Agent Stop? Evidence-Carrying Termination for Tool-Using LLMs

智能体何时可以停止？带证据的工具使用大语言模型终止机制

Jason Liu

机构 * University of California San Diego（加利福尼亚大学圣迭戈分校）

专题命中其他LLM ：LLM(abstract)；分类 cs.AI、cs.LG

AI总结该研究针对工具使用大语言模型的终止问题，提出带证据的终止（ECT）机制，经实验验证其能显著减少不安全完成与过早无支持终止，满足非劣效性要求，可实现成功恢复。

2608.22695 2026-08-25 cs.CL cs.AI cs.IR 新提交 62%

Enrich-Retrieve-Rank: Scaling Capability Discovery Beyond In-Context Routing

Enrich-Retrieve-Rank：将能力发现扩展至上下文路由之外

Nazib Sorathiya, Daniel Zhang, Bardiya Akhbari

机构 * Amazon AGI（亚马逊AGI）

专题命中其他LLM ：LLM(abstract)；分类 cs.CL、cs.AI

AI总结该研究提出Enrich-Retrieve-Rank流程，将能力发现从上下文路由扩展，经实验验证其在大规模MATS组件场景下，比Full-Ctx、Search&Pick基线性能更优且成本更低，已作为多智能体平台的默认能力发现层投入生产。

Comments 11 pages, 4 figures, and 12 tables

2608.16971 2026-08-19 cs.AI cs.LG 新提交 62%

FedPref: Federated Preference Learning for Structured Radiology Report Extraction

FedPref：用于结构化放射报告抽取的联邦偏好学习

Flint Xiaofeng Fan, Cheston Tan, Yew-Soon Ong, Roger Wattenhofer

机构 * ETH Zurich（苏黎世联邦理工学院）； Agency for Science, Technology and Research (A*STAR)（新加坡科技研究局）

专题命中其他LLM ：language model(abstract)；分类 cs.AI、cs.LG

AI总结 FedPref 是用于结构化放射报告抽取的联邦偏好学习方法，通过冻结公共语言模型、本地排序与共享模型更新训练适配器，在不均数据场景下提升抽取性能且无需共享敏感数据。

Comments Accepted at ELAMI 2026, held in conjunction with MICCAI 2026. To appear in the Springer proceedings

1. 其他LLM 12287 篇

Taming the Beast: Learning to Control Neural Conversational Models

ReCAM@IITK at SemEval-2021 Task 4: BERT and ALBERT based Ensemble for Abstract Word Prediction

Comparing Approaches to Dravidian Language Identification

Adv-OLM: Generating Textual Adversaries via OLM

AutoDropout: Learning Dropout Patterns to Regularize Deep Networks

Directed Beam Search: Plug-and-Play Lexically Constrained Language Generation

Recurrent Point Review Models

Compositional Demographic Word Embeddings

Vokenization: Improving Language Understanding with Contextualized, Visual-Grounded Supervision

GenAug: Data Augmentation for Finetuning Text Generators

A Systematic Characterization of Sampling Algorithms for Open-ended Language Generation

Multi-agent Communication meets Natural Language: Synergies between Functional and Structural Language Learning

Social Biases in NLP Models as Barriers for Persons with Disabilities

Explaining Question Answering Models through Text Generation

Cyclical Annealing Schedule: A Simple Approach to Mitigating KL Vanishing

GLTR: Statistical Detection and Visualization of Generated Text

Conditional BERT Contextual Augmentation

Learning Visual N-Grams from Web Data

Quasi-Recurrent Neural Networks

End-to-End Attention-based Large Vocabulary Speech Recognition

ANX: Protocol-First Design for AI Agent Interaction with a Supporting 3EX Decoupled Architecture

Child vs. machine language learning: Can the logical structure of human language unleash LLMs?

Wiki-LLaVA: Hierarchical Retrieval-Augmented Generation for Multimodal LLMs

On the Security Vulnerabilities of Text-to-SQL Models

Grocery to General Merchandise: A Cross-Pollination Recommender using LLMs and Real-Time Cart Context

Squeezing More from Limited Data with Recursive Transformers

A Single Suffix to Break Them All: Basin-Aware Jailbreaks for Merged Model Families

When May an Agent Stop? Evidence-Carrying Termination for Tool-Using LLMs

Enrich-Retrieve-Rank: Scaling Capability Discovery Beyond In-Context Routing

FedPref: Federated Preference Learning for Structured Radiology Report Extraction