arXivDaily arXiv每日学术速递 周一至周五更新

AI 大模型

语言大模型 / LLM

大语言模型、预训练、指令微调、后训练和语言模型应用。

共收录 11654 信号源:cs.CL, cs.AI, cs.LG

1. 指令微调 11654 篇

2512.07374 2025-12-09 cs.LG cs.CL 81%

Recover-to-Forget: Gradient Reconstruction from LoRA for Efficient LLM Unlearning

恢复-遗忘:从LoRA恢复梯度以实现高效的LLM反学习

Yezi Liu, Hanning Chen, Wenjun Huang, Yang Ni, Mohsen Imani

机构 * University of California, Irvine(加州大学尔湾分校) Purdue University Northwest(普渡大学西北分校)

专题命中 指令微调 :LLM(title);foundation model(abstract);分类 cs.CL、cs.LG

AI总结 R2F通过从LoRA适配器更新中恢复梯度方向,实现高效LLM反学习,无需完整微调或内部参数访问。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.06811 2025-12-09 cs.CV cs.AI cs.LG cs.MM 81%

RMAdapter: Reconstruction-based Multi-Modal Adapter for Vision-Language Models

RMAdapter: 基于重建的多模态适配器用于视觉-语言模型

Xiang Lin, Weixin Li, Shu Guo, Lihong Wang, Di Huang

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

AI总结 RMAdapter通过双分支架构平衡通用与任务特定知识,提升视觉-语言模型在多模态迁移学习中的性能。

Comments Accepted by AAAI 2026(Oral)

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.04044 2025-12-04 cs.LG cs.AI cs.CR 81%

MarkTune: Improving the Quality-Detectability Trade-off in Open-Weight LLM Watermarking

MarkTune: 改善开放权重语言模型水印中的质量-可检测性权衡

Yizhou Zhao, Zhiwei Steven Wu, Adam Block

机构 * University of Pennsylvania(宾夕法尼亚大学) Carnegie Mellon University(卡内基梅隆大学) Columbia University(哥伦比亚大学)

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

AI总结 MarkTune通过理论框架提升开放权重语言模型中质量与可检测性的平衡,优于GaussMark并保持生成质量。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.03499 2025-12-04 cs.CV cs.AI cs.CL 81%

NAS-LoRA: Empowering Parameter-Efficient Fine-Tuning for Visual Foundation Models with Searchable Adaptation

NAS-LoRA: 通过可搜索适应增强视觉基础模型的参数高效微调

Renqi Chen, Haoyang Su, Shixiang Tang

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

AI总结 NAS-LoRA 通过引入可搜索适应的神经网络架构搜索块,提升视觉基础模型在特定领域中的参数高效微调性能,减少训练成本24.14%。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.03463 2025-12-04 cs.CV cs.AI cs.CL 81%

Text-Printed Image: Bridging the Image-Text Modality Gap for Text-centric Training of Large Vision-Language Models

文本打印图像:为以文本为中心的大型视觉-语言模型训练弥合图像-文本模态差距

Shojiro Yamabe, Futa Waseda, Daiki Shiono, Tsubasa Takahashi

机构 * Turing Inc.(图灵公司) Institute of Science Tokyo(东京科学研究院) The University of Tokyo(东京大学) Tohoku University(东北大学)

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

AI总结 本研究提出文本打印图像(TPI)技术,通过生成合成图像弥合图像-文本模态差距,提升以文本为中心的大型视觉-语言模型训练效果。

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.09809 2025-12-03 cs.CV cs.AI cs.LG 81%

Test-Time Spectrum-Aware Latent Steering for Zero-Shot Generalization in Vision-Language Models

测试时谱感知的潜在引导用于视觉-语言模型中的零样本泛化

Konstantinos M. Dafnis, Dimitris N. Metaxas

机构 * Rutgers University(罗格斯大学)

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

AI总结 本文提出STP,一种轻量级的测试时适应框架,通过谱感知方法引导潜在表示,提升视觉-语言模型在零样本泛化中的性能,同时保持高效推理速度和低内存消耗。

Comments NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.01038 2025-12-02 cs.LG cs.AI 81%

FMTK: A Modular Toolkit for Composable Time Series Foundation Model Pipelines

FMTK:用于可组合时间序列基础模型流水线的模块化工具包

Hetvi Shastri, Pragya Sharma, Walid A. Hanafy, Mani Srivastava, Prashant Shenoy

机构 * University of Massachusetts Amherst(马萨诸塞大学阿默斯特分校) University of California Los Angeles(加州大学洛杉矶分校)

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

AI总结 FMTK提供了一个模块化工具包,用于构建和微调时间序列基础模型流水线,通过标准化的主干和组件抽象实现灵活组合和高效性能。

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.01196 2025-11-26 cs.LG cs.AI cs.ET cs.HC 81%

Are Large Brainwave Foundation Models Capable Yet? Insights from Fine-tuning

大规模脑电基础模型是否已具备能力?来自微调的洞察

Na Lee, Konstantinos Barmpas, Yannis Panagakis, Dimitrios Adamos, Nikolaos Laskaris, Stefanos Zafeiriou

机构 * Imperial College London(伦敦帝国学院) Archimedes / Athena Research Unit(阿基米德/雅典娜研究单位) Aristotle University of Thessaloniki(雅典娜大学) Kapodistrian University of Athens(雅典kapodistrian大学)

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

AI总结 本文通过微调实验评估了大规模脑电基础模型的能力,发现其在BCI任务中效率有限,提出LoRA技术可提升性能,强调需重新设计架构以提升脑电分析效果。

Journal ref International Conference on Machine Learning (ICML) 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2507.16887 2025-11-25 cs.CR cs.AI cs.LG cs.SE 81%

Revisiting Pre-trained Language Models for Vulnerability Detection

重新审视预训练语言模型用于漏洞检测

Youpeng Li, Weiliang Qi, Xuyu Wang, Fuxun Yu, Xinda Wang

机构 * University of Texas at Dallas(德克萨斯大学达拉斯分校) Florida International University(佛罗里达国际大学) Microsoft(微软公司)

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

AI总结 本文重新审视预训练语言模型在漏洞检测中的应用,发现特定设计的模型在性能上更优,但面临现实场景中的挑战,如复杂依赖检测和标注错误问题。

Comments Accepted by the 21st ACM ASIA Conference on Computer and Communications Security (AsiaCCS 2026)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.11690 2025-11-18 cs.LG cs.AI cs.CV 81%

Doubly Debiased Test-Time Prompt Tuning for Vision-Language Models

Fei Song, Yi Li, Rui Wang, Jiahuan Zhou, Changwen Zheng, Jiangmeng Li

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

Comments Accepted by AAAI2026

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.12684 2025-11-17 cs.LG cs.AI cs.DB cs.SI 81%

Towards Effective Federated Graph Foundation Model via Mitigating Knowledge Entanglement

Yinlin Zhu, Xunkai Li, Jishuo Jia, Miao Hu, Di Wu, Meikang Qiu

机构 * Sun Yat-sen University(中山大学) Beijing Institute of Technology(北京理工大学) Shandong University(山东大学) Augusta University(奥古斯塔大学)

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

Comments Accepted by NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.03149 2025-11-06 cs.LG cs.AI 81%

Forecast2Anomaly (F2A): Adapting Multivariate Time Series Foundation Models for Anomaly Prediction

Atif Hassan, Tarun Kumar, Ashish Mishra, Sergey Serebryakov, Satish Kumar Mopur, Phanidhar Koganti, Murthy Chelankuri, Ramanagopal Vogety, Suparna Bhattacharya, Martin Foltin

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.03128 2025-11-06 cs.LG cs.CL 81%

From Insight to Exploit: Leveraging LLM Collaboration for Adaptive Adversarial Text Generation

Najrin Sultana, Md Rafi Ur Rashid, Kang Gu, Shagufta Mehnaz

机构 * The Pennsylvania State University(宾夕法尼亚州立大学) Dartmouth College(达特茅斯学院)

专题命中 指令微调 :LLM(title,abstract);分类 cs.CL、cs.LG

Comments Findings of the Association for Computational Linguistics: EMNLP 2025 (camera-ready)

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.09707 2025-11-06 cs.LG cs.AI cs.CV 81%

Revisiting semi-supervised learning in the era of foundation models

Ping Zhang, Zheda Mai, Quang-Huy Nguyen, Wei-Lun Chao

机构 * The Ohio State University(俄亥俄州立大学)

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

Comments The paper has been accepted to NeurIPS 2025. Ping Zhang and Zheda Mai contributed equally to this work

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.12740 2025-11-05 cs.CL cs.AI 81%

Hey, wait a minute: on at-issue sensitivity in Language Models

Sanghee J. Kim, Kanishka Misra

机构 * The University of Chicago(芝加哥大学) The University of Texas at Austin(德克萨斯大学奥斯汀分校)

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

Comments 10 pages, 5 figures, 3 tables. See https://github.com/sangheek16/hey-wait-a-minute for code and data

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.23502 2025-10-29 cs.CV cs.AI cs.LG cs.RO 81%

Boosting Omnidirectional Stereo Matching with a Pre-trained Depth Foundation Model

Jannik Endres, Oliver Hahn, Charles Corbière, Simone Schaub-Meyer, Stefan Roth, Alexandre Alahi

机构 * École Polytechnique Fédérale de Lausanne (EPFL)(瑞士联邦理工学院) Technical University of Darmstadt, Department of Computer Science(德累斯顿技术大学计算机科学系)

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

Comments Accepted at IROS 2025. Project page: https://vita-epfl.github.io/DFI-OmniStereo-website/

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.22301 2025-10-28 cs.LG cs.AI 81%

AnyECG-Lab: An Exploration Study of Fine-tuning an ECG Foundation Model to Estimate Laboratory Values from Single-Lead ECG Signals

Yujie Xiao, Gongzhen Tang, Wenhui Liu, Jun Li, Guangkun Nie, Zhuoran Kan, Deyun Zhang, Qinghao Zhao, Shenda Hong

机构 * Institute of Medical Technology, Peking University Health Science Center(北京大学人民医院医学技术研究所) National Institute of Health Data Science, Peking University(北京大学国家健康数据科学研究院) School of Intelligence Science and Technology, Peking University(北京大学智能科学与技术学院) HeartVoice Medical Technology(心声医疗技术) Department of Cardiology, Peking University People’s Hospital(北京大学人民医院心内科) Institute for Artificial Intelligence, Peking University(北京大学人工智能研究院) State Key Laboratory of Vascular Homeostasis and Remodeling, NHC Key Laboratory of Cardiovascular Molecular Biology and Regulatory Peptides, Peking University(国家心血管病分子生物学与调节肽重点实验室,北京大学)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.20990 2025-10-27 cs.LG cs.CL cs.CV 81%

SharpZO: Hybrid Sharpness-Aware Vision Language Model Prompt Tuning via Forward-Only Passes

Yifan Yang, Zhen Zhang, Rupak Vignesh Swaminathan, Jing Liu, Nathan Susanj, Zheng Zhang

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.20342 2025-10-24 cs.CL cs.AI 81%

Teaching Language Models to Reason with Tools

Chengpeng Li, Zhengyang Tang, Ziniu Li, Mingfeng Xue, Keqin Bao, Tian Ding, Ruoyu Sun, Benyou Wang, Xiang Wang, Junyang Lin, Dayiheng Liu

机构 * University of Science and Technology of China(中国科学技术大学) Alibaba Inc.(阿里巴巴公司) The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳)) Shenzhen International Center for Industrial and Applied Mathematics(深圳国际工业与应用数学中心) Shenzhen Research Institute of Big Data(深圳大数据研究院)

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

Comments NIPS2025 Accepted

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.20222 2025-10-24 cs.LG cs.AI 81%

QKCV Attention: Enhancing Time Series Forecasting with Static Categorical Embeddings for Both Lightweight and Pre-trained Foundation Models

Hao Wang, Baojun Ma

机构 * Independent Researcher(独立研究者) Key Laboratory of Brain-Machine Intelligence for Information Behavior (Ministry of Education and Shanghai)(信息行为脑机智能关键实验室) School of Business and Management(商学院)

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

Comments 10 pages, 5 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.00237 2025-10-02 cs.LG cs.AI 81%

Debunk the Myth of SFT Generalization

Xiaofeng Lin, Hejian Sang, Zhipeng Wang, Xuezhou Zhang

机构 * Boston University(波士顿大学) LinkedIn

专题命中 指令微调 :SFT(title,abstract);分类 cs.AI、cs.LG

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.26598 2025-10-01 cs.CR cs.AI cs.LG 81%

Are Robust LLM Fingerprints Adversarially Robust?

Anshul Nasery, Edoardo Contente, Alkin Kaz, Pramod Viswanath, Sewoong Oh

机构 * University of Washington(华盛顿大学) Princeton University(普林斯顿大学) Sentient

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.25274 2025-10-01 q-bio.GN cs.AI cs.LG 81%

DNABERT-2: Fine-Tuning a Genomic Language Model for Colorectal Gene Enhancer Classification

Darren King, Yaser Atlasi, Gholamreza Rafiee

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

Comments 10 pages, 10 figures, 2 tables

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.20230 2025-10-01 cs.LG cs.AI 81%

Beyond Sharp Minima: Robust LLM Unlearning via Feedback-Guided Multi-Point Optimization

Wenhan Wu, Zheyuan Liu, Chongyang Gao, Ren Wang, Kaize Ding

机构 * Department of Statistics and Data Science(统计与数据科学系) Northwestern University(西北大学) Department of Computer Science and Engineering(计算机科学与工程系) University of Notre Dame(圣母大学) Department of Computer Science(计算机科学系) Illinois Institute of Technology(伊利诺伊理工学院)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.22906 2025-09-30 cs.CL cs.AI 81%

Extract-0: A Specialized Language Model for Document Information Extraction

Henrique Godoy

机构 * Inteli São Paulo(Inteli圣保罗)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.20107 2025-09-26 cs.CV cs.AI cs.LG cs.RO 81%

Hyperspectral Adapter for Semantic Segmentation with Vision Foundation Models

Juana Valeria Hurtado, Rohit Mohan, Abhinav Valada

机构 * Department of Computer Science, University of Freiburg(弗赖堡大学计算机科学系) Baden-Württemberg Stiftung gGmbH(巴登-符腾堡基金会) Bosch Research(博世研究)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.13448 2025-09-23 cs.CL cs.AI 81%

CIE: Controlling Language Model Text Generations Using Continuous Signals

Vinay Samuel, Harshita Diddee, Yiming Zhang, Daphne Ippolito

机构 * University of Maryland, College Park(马里兰大学 College Park分校) Carnegie Mellon University(卡内基梅隆大学)

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

Comments EMNLP Main 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2402.15449 2025-09-09 cs.CL cs.LG 81%

Repetition Improves Language Model Embeddings

Jacob Mitchell Springer, Suhas Kotha, Daniel Fried, Graham Neubig, Aditi Raghunathan

机构 * Carnegie Mellon University(卡内基梅隆大学)

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

Comments ICLR 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.04473 2025-09-08 cs.CL cs.AI 81%

SpeechLLM: Unified Speech and Language Model for Enhanced Multi-Task Understanding in Low Resource Settings

Jaekwon Yoo, Kunal Chandiramani, Divya Tadimeti, Abenezer Girma, Chandra Dhir

机构 * JPMorganChaseUSA(摩根大通公司) Columbia UniversityUSA(哥伦比亚大学)

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

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.19609 2025-08-28 cs.LG cs.AI q-fin.CP 81%

FinCast: A Foundation Model for Financial Time-Series Forecasting

Zhuohang Zhu, Haodong Chen, Qiang Qu, Vera Chung

机构 * School of Computer Science(计算机科学系) The University of Sydney(悉尼大学)

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

详情

展开后加载摘要…

URL PDF HTML 收藏