语言大模型 / LLM - arXivDaily 专题

2511.17053 2025-11-24 cs.CV cs.AI 83%

OmniPT: Unleashing the Potential of Large Vision Language Models for Pedestrian Tracking and Understanding

OmniPT：释放大型视觉语言模型在行人跟踪与理解中的潜力

Teng Fu, Mengyang Zhao, Ke Niu, Kaixin Peng, Bin Li

专题命中指令微调：language model(title)；foundation model(abstract)；SFT(abstract)；分类 cs.AI

AI总结 OmniPT 通过结合强化学习和微调方法，提升大型视觉语言模型在行人跟踪和语义理解中的性能。

Comments AAAI 2026

2511.16709 2025-11-24 cs.CR cs.AI 83%

AutoBackdoor: Automating Backdoor Attacks via LLM Agents

AutoBackdoor: 通过LLM代理自动化后门攻击

Yige Li, Zhe Li, Wei Zhao, Nay Myat Min, Hanxun Huang, Xingjun Ma, Jun Sun

机构 * Singapore Management University（新加坡管理大学）； The University of Melbourne（墨尔本大学）； Fudan University（复旦大学）

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.AI

AI总结 AutoBackdoor通过LLM代理自动化后门攻击，实现高效触发器生成和毒化数据构建，验证了对抗防御的脆弱性。

Comments 23 pages

2511.07991 2025-11-19 cs.AI 83%

VSPO: Validating Semantic Pitfalls in Ontology via LLM-Based CQ Generation

Hyojun Choi, Seokju Hwang, Kyong-Ho Lee

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.AI

Comments Accepted at AAAI 2026 oral

2503.17352 2025-11-12 cs.CV cs.CL 83%

OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles

Yihe Deng, Hritik Bansal, Fan Yin, Nanyun Peng, Wei Wang, Kai-Wei Chang

专题命中指令微调：SFT(title,abstract)；language model(abstract)；分类 cs.CL

Comments 23 pages, 11 figures, 8 tables

2508.18253 2025-11-04 cs.CL 83%

From BERT to LLMs: Comparing and Understanding Chinese Classifier Prediction in Language Models

Ziqi Zhang, Jianfei Ma, Emmanuele Chersoni, Jieshun You, Zhaoxin Feng

机构 * Department of Language Science and Technology, The Hong Kong Polytechnic University（语言科学与技术系，香港理工大学）

专题命中指令微调：language model(title,abstract)；large language model(abstract)；分类 cs.CL

2502.01586 2025-10-28 cs.LG 83%

SubTrack++ : Gradient Subspace Tracking for Scalable LLM Training

Sahar Rajabi, Nayeema Nonta, Sirisha Rambhatla

机构 * Critical ML, Department of Management Science and Engineering, University of Waterloo（Critical ML，管理科学与工程系，滑铁卢大学）

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.LG

2510.20033 2025-10-24 cs.CL 83%

Improving Transfer Learning for Sequence Labeling Tasks by Adapting Pre-trained Neural Language Models

David Dukić

机构 * FACULTY OF ELECTRICAL ENGINEERING AND COMPUTING（电气工程与计算学院）； University of Zagreb, Faculty of Electrical Engineering and Computing（Zagreb大学电气工程与计算学院）； Department of Electronics, Microelectronics, Computer and Intelligent systems（电子、微电子、计算机和智能系统系）； Text Analysis and Knowledge Engineering Laboratory (TakeLab)（文本分析与知识工程实验室）

专题命中指令微调：language model(title,abstract)；large language model(abstract)；分类 cs.CL

2510.16022 2025-10-21 cs.LG cs.SE 83%

Breaking Memorization Barriers in LLM Code Fine-Tuning via Information Bottleneck for Improved Generalization

Changsheng Wang, Xin Chen, Sijia Liu, Ke Ding

机构 * The OPTML Lab, Michigan State University（密歇根州立大学OPTML实验室）； Intel（英特尔）； IBM Research（IBM研究）

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.LG

2510.11408 2025-10-21 cs.CL 83%

Valid Survey Simulations with Limited Human Data: The Roles of Prompting, Fine-Tuning, and Rectification

Stefan Krsteski, Giuseppe Russo, Serina Chang, Robert West, Kristina Gligorić

机构 * EPFL（苏黎世联邦理工学院）； Stanford University（斯坦福大学）； University of California, Berkeley（加州大学伯克利分校）； Johns Hopkins University（约翰霍普金斯大学）

专题命中指令微调：prompting(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments 19 pages, 4 figures, 9 tables

2510.14470 2025-10-17 cs.CR cs.AI 83%

Stealthy Dual-Trigger Backdoors: Attacking Prompt Tuning in LM-Empowered Graph Foundation Models

Xiaoyu Xue, Yuni Lai, Chenxi Huang, Yulin Zhu, Gaolei Li, Xiaoge Zhang, Kai Zhou

机构 * HKPolyU（香港PolyU）； Hong Kong Chu Hai College（香港 Chu Hai 学院）； Shanghai Jiao Tong University（上海交通大学）

专题命中指令微调：foundation model(title,abstract)；language model(abstract)；分类 cs.AI

2510.14453 2025-10-17 cs.CL 83%

Natural Language Tools: A Natural Language Approach to Tool Calling In Large Language Agents

Reid T. Johnson, Michelle D. Pain, Jordan D. West

专题命中指令微调：language agent(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments 31 pages, 7 figures

2510.10517 2025-10-14 cs.PL cs.AI cs.SE 83%

ECO: Enhanced Code Optimization via Performance-Aware Prompting for Code-LLMs

Su-Hyeon Kim, Joonghyuk Hahn, Sooyoung Cha, Yo-Sub Han

专题命中指令微调：prompting(title,abstract)；LLM(abstract)；分类 cs.AI

2510.06354 2025-10-09 cs.CL 83%

LLM Bias Detection and Mitigation through the Lens of Desired Distributions

Ingroj Shrestha, Padmini Srinivasan

机构 * University of Iowa（爱荷华大学）

专题命中指令微调：LLM(title,abstract)；language model(abstract)；分类 cs.CL

Comments Accepted to EMNLP 2025

2503.20752 2025-10-07 cs.CV cs.AI 83%

Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models

Huajie Tan, Yuheng Ji, Xiaoshuai Hao, Xiansheng Chen, Pengwei Wang, Zhongyuan Wang, Shanghang Zhang

机构 * State Key Laboratory of Multimedia Information Processing, School of Computer Science, Peking University（多媒体信息处理国家重点实验室，计算机科学学院，北京大学）； Beijing Academy of Artificial Intelligence（北京人工智能研究院）； Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）； School of Artificial Intelligence, University of Chinese Academy of Sciences（中国科学院自动化研究所人工智能学院）

专题命中指令微调：language model(title,abstract)；SFT(abstract)；分类 cs.AI

Comments 51 pages, 23 figures, NeurIPS'25

2510.00088 2025-10-02 cs.AI cs.CY 83%

Judging by Appearances? Auditing and Intervening Vision-Language Models for Bail Prediction

Sagnik Basu, Shubham Prakash, Ashish Maruti Barge, Siddharth D Jaiswal, Abhisek Dash, Saptarshi Ghosh, Animesh Mukherjee

机构 * Indian Institute of Technology Kharagpur（印度理工学院Khargapur分校）； Max Planck Institute for Software Systems（马克斯·普朗克软件系统研究所）

专题命中指令微调：language model(title,abstract)；large language model(abstract)；分类 cs.AI

2506.01702 2025-09-25 cs.CL 83%

mdok of KInIT: Robustly Fine-tuned LLM for Binary and Multiclass AI-Generated Text Detection

Dominik Macko

机构 * Kempelen Institute of Intelligent Technologies（凯普伦智能技术研究所）

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments 1st rank in both subtasks of the Voight-Kampff Generative AI Detection 2025 shared task (PAN@CLEF 2025)

Journal ref CLEF 2025 Working Notes

2509.16660 2025-09-23 cs.CL 83%

Redefining Experts: Interpretable Decomposition of Language Models for Toxicity Mitigation

Zuhair Hasan Shaik, Abdullah Mazhar, Aseem Srivastava, Md Shad Akhtar

机构 * IIIT Dharwad, India（印度IIIT达尔瓦德大学）； IIIT Delhi, India（印度IIIT德里大学）； FLaME-NLP Lab, IIIT Delhi（IIIT德里大学FLaME-NLP实验室）

专题命中指令微调：language model(title,abstract)；large language model(abstract)；分类 cs.CL

Comments Accepted to the NeurIPS 2025 Research Track

2412.00251 2025-09-23 cs.AI cs.HC 83%

Fine-Tuning Open-Weight Language Models to Deliver Cognitive Behavioral Therapy for Depression: A Feasibility Study

Talha Tahir

机构 * Department of Psychiatry University of Toronto（精神病学系多伦多大学）

专题命中指令微调：language model(title,abstract)；large language model(abstract)；分类 cs.AI

2509.13906 2025-09-18 cs.LG 83%

TFMAdapter: Lightweight Instance-Level Adaptation of Foundation Models for Forecasting with Covariates

Afrin Dange, Sunita Sarawagi

机构 * Indian Institute of Technology Bombay Centre for Machine Intelligence（印度理工学院班加罗尔中心人工智能与数据科学中心）； Indian Institute of Technology Bombay Department of Computer Science（印度理工学院班加罗尔计算机科学与工程系）

专题命中指令微调：foundation model(title,abstract)；pretraining(abstract)；分类 cs.LG

Comments Accepted at CIKM 2025

2509.02915 2025-09-04 cs.CL 83%

English Pronunciation Evaluation without Complex Joint Training: LoRA Fine-tuned Speech Multimodal LLM

Taekyung Ahn, Hosung Nam

机构 * Enuma, Inc.（Enuma公司）； Korea University（韩国大学）

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

2509.01142 2025-09-03 cs.CL 83%

Dream-Coder 7B: An Open Diffusion Language Model for Code

Zhihui Xie, Jiacheng Ye, Lin Zheng, Jiahui Gao, Jingwei Dong, Zirui Wu, Xueliang Zhao, Shansan Gong, Xin Jiang, Zhenguo Li, Lingpeng Kong

机构 * The University of Hong Kong（香港大学）； Huawei Noah’s Ark Lab（华为诺亚实验室）

专题命中指令微调：language model(title,abstract)；post-training(abstract)；分类 cs.CL

2504.11195 2025-08-28 cs.LG cs.CR cs.CV 83%

R-TPT: Improving Adversarial Robustness of Vision-Language Models through Test-Time Prompt Tuning

Lijun Sheng, Jian Liang, Zilei Wang, Ran He

机构 * University of Science and Technology of China（中国科学技术大学）； NLPR & MAIS, Institute of Automation, Chinese Academy of Sciences（中国科学院自动化研究所）； University of Chinese Academy of Sciences（中国科学院大学）

专题命中指令微调：language model(title,abstract)；foundation model(abstract)；分类 cs.LG

Comments CVPR 2025 (Corrected the results on the Aircraft dataset)

2508.17158 2025-08-26 cs.LG 83%

Towards Safeguarding LLM Fine-tuning APIs against Cipher Attacks

Jack Youstra, Mohammed Mahfoud, Yang Yan, Henry Sleight, Ethan Perez, Mrinank Sharma

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.LG

2508.06360 2025-08-25 cs.CL 83%

Cyberbullying Detection via Aggression-Enhanced Prompting

Aisha Saeid, Anu Sabu, Girish A. Koushik, Ferrante Neri, Diptesh Kanojia

机构 * NICE Research Group & Institute for People-Centred AI, School of Computer Science & Electronic Engineering, University of Surrey, UK（NICE研究组及以人为本的人工智能研究所、计算机科学与电子工程学院、萨里大学）

专题命中指令微调：prompting(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments Accepted to RANLP 2025

2503.12530 2025-08-20 cs.CL 83%

Basic Category Usage in Vision Language Models

Hunter Sawyer, Jesse Roberts, Kyle Moore

机构 * Computer Science, Tennessee Tech University（田纳西科技大学计算机科学系）； Computer Science, Vanderbilt University（范德比大学计算机科学系）

专题命中指令微调：language model(title,abstract)；prompting(abstract)；分类 cs.CL

2506.00582 2025-07-29 cs.AI 83%

Do Language Models Mirror Human Confidence? Exploring Psychological Insights to Address Overconfidence in LLMs

Chenjun Xu, Bingbing Wen, Bin Han, Robert Wolfe, Lucy Lu Wang, Bill Howe

机构 * University of Washington（华盛顿大学）； Allen Institute for AI（人工智能研究所）

专题命中指令微调：language model(title)；LLM(abstract)；prompting(abstract)；分类 cs.AI

Comments Accepted by ACL 2025 Findings, 20 pages

2502.20292 2025-07-25 cs.CV cs.LG 83%

Visual Adaptive Prompting for Compositional Zero-Shot Learning

Kyle Stein, Arash Mahyari, Guillermo Francia, Eman El-Sheikh

机构 * University of West Florida（乌斯托尔大学）； Florida Institute for Human and Machine Cognition (IHMC)（佛罗里达人类与机器认知研究所）

专题命中指令微调：prompting(title,abstract)；language model(abstract)；分类 cs.LG

2507.14887 2025-07-22 cs.CL 83%

MEKiT: Multi-source Heterogeneous Knowledge Injection Method via Instruction Tuning for Emotion-Cause Pair Extraction

Shiyi Mu, Yongkang Liu, Shi Feng, Xiaocui Yang, Daling Wang, Yifei Zhang

机构 * Northeastern University（东北大学）

专题命中指令微调：instruction tuning(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

Comments Accepted by CogSci

2507.04889 2025-07-08 cs.LG 83%

Fine-tuning on simulated data outperforms prompting for agent tone of voice

Ingo Marquardt, Philippe Brule

机构 * Restack

专题命中指令微调：prompting(title,abstract)；language model(abstract)；分类 cs.LG

Comments 22 pages, 5 figures, 6 tables

2507.00601 2025-07-03 cs.CL 83%

Transferable Modeling Strategies for Low-Resource LLM Tasks: A Prompt and Alignment-Based Approach

Shuangquan Lyu, Yingnan Deng, Guiran Liu, Zhen Qi, Ruotong Wang

专题命中指令微调：LLM(title)；large language model(abstract)；language model(abstract)；分类 cs.CL

1. 指令微调 11601 篇

OmniPT: Unleashing the Potential of Large Vision Language Models for Pedestrian Tracking and Understanding

AutoBackdoor: Automating Backdoor Attacks via LLM Agents

VSPO: Validating Semantic Pitfalls in Ontology via LLM-Based CQ Generation

OpenVLThinker: Complex Vision-Language Reasoning via Iterative SFT-RL Cycles

From BERT to LLMs: Comparing and Understanding Chinese Classifier Prediction in Language Models

SubTrack++ : Gradient Subspace Tracking for Scalable LLM Training

Improving Transfer Learning for Sequence Labeling Tasks by Adapting Pre-trained Neural Language Models

Breaking Memorization Barriers in LLM Code Fine-Tuning via Information Bottleneck for Improved Generalization

Valid Survey Simulations with Limited Human Data: The Roles of Prompting, Fine-Tuning, and Rectification

Stealthy Dual-Trigger Backdoors: Attacking Prompt Tuning in LM-Empowered Graph Foundation Models

Natural Language Tools: A Natural Language Approach to Tool Calling In Large Language Agents

ECO: Enhanced Code Optimization via Performance-Aware Prompting for Code-LLMs

LLM Bias Detection and Mitigation through the Lens of Desired Distributions

Reason-RFT: Reinforcement Fine-Tuning for Visual Reasoning of Vision Language Models

Judging by Appearances? Auditing and Intervening Vision-Language Models for Bail Prediction

mdok of KInIT: Robustly Fine-tuned LLM for Binary and Multiclass AI-Generated Text Detection

Redefining Experts: Interpretable Decomposition of Language Models for Toxicity Mitigation

Fine-Tuning Open-Weight Language Models to Deliver Cognitive Behavioral Therapy for Depression: A Feasibility Study

TFMAdapter: Lightweight Instance-Level Adaptation of Foundation Models for Forecasting with Covariates

English Pronunciation Evaluation without Complex Joint Training: LoRA Fine-tuned Speech Multimodal LLM

Dream-Coder 7B: An Open Diffusion Language Model for Code

R-TPT: Improving Adversarial Robustness of Vision-Language Models through Test-Time Prompt Tuning

Towards Safeguarding LLM Fine-tuning APIs against Cipher Attacks

Cyberbullying Detection via Aggression-Enhanced Prompting

Basic Category Usage in Vision Language Models

Do Language Models Mirror Human Confidence? Exploring Psychological Insights to Address Overconfidence in LLMs

Visual Adaptive Prompting for Compositional Zero-Shot Learning

MEKiT: Multi-source Heterogeneous Knowledge Injection Method via Instruction Tuning for Emotion-Cause Pair Extraction

Fine-tuning on simulated data outperforms prompting for agent tone of voice

Transferable Modeling Strategies for Low-Resource LLM Tasks: A Prompt and Alignment-Based Approach