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 7583 信号源:cs.CL, cs.AI, cs.LG

1. 知识编辑与模型理解 7583 篇

2405.13951 2024-05-24 cs.CV 78%

Text Prompting for Multi-Concept Video Customization by Autoregressive Generation

Divya Kothandaraman, Kihyuk Sohn, Ruben Villegas, Paul Voigtlaender, Dinesh Manocha, Mohammad Babaeizadeh

专题命中 知识编辑与模型理解 :prompting(title,abstract)

Comments Paper accepted to AI4CC Workshop at CVPR 2024

详情

展开后加载摘要…

URL PDF HTML 收藏
2404.03592 2024-05-24 cs.CL cs.AI cs.LG 78%

ReFT: Representation Finetuning for Language Models

Zhengxuan Wu, Aryaman Arora, Zheng Wang, Atticus Geiger, Dan Jurafsky, Christopher D. Manning, Christopher Potts

专题命中 知识编辑与模型理解 :language model(title);分类 cs.CL、cs.AI、cs.LG

Comments preprint

详情

展开后加载摘要…

URL PDF HTML 收藏
2404.11683 2024-04-19 cs.RO cs.CV 78%

Unifying Scene Representation and Hand-Eye Calibration with 3D Foundation Models

Weiming Zhi, Haozhan Tang, Tianyi Zhang, Matthew Johnson-Roberson

专题命中 知识编辑与模型理解 :foundation model(title,abstract)

详情

展开后加载摘要…

URL PDF HTML 收藏
2404.10193 2024-04-17 cs.CV 78%

Consistency and Uncertainty: Identifying Unreliable Responses From Black-Box Vision-Language Models for Selective Visual Question Answering

Zaid Khan, Yun Fu

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments CVPR 2024

详情

展开后加载摘要…

URL PDF HTML 收藏
2402.07329 2024-04-17 cs.CV 78%

The Bias of Harmful Label Associations in Vision-Language Models

Caner Hazirbas, Alicia Sun, Yonathan Efroni, Mark Ibrahim

专题命中 知识编辑与模型理解 :language model(title,abstract)

详情

展开后加载摘要…

URL PDF HTML 收藏
2403.12693 2024-03-20 cs.CV 78%

As Firm As Their Foundations: Can open-sourced foundation models be used to create adversarial examples for downstream tasks?

Anjun Hu, Jindong Gu, Francesco Pinto, Konstantinos Kamnitsas, Philip Torr

专题命中 知识编辑与模型理解 :foundation model(title,abstract)

详情

展开后加载摘要…

URL PDF HTML 收藏
2311.14339 2024-03-07 cs.CV 78%

Towards Concept-based Interpretability of Skin Lesion Diagnosis using Vision-Language Models

Cristiano Patrício, Luís F. Teixeira, João C. Neves

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments Accepted for publication in IEEE ISBI 2024

详情

展开后加载摘要…

URL PDF HTML 收藏
2401.12181 2024-01-23 cs.LG cs.AI cs.CL 78%

Universal Neurons in GPT2 Language Models

Wes Gurnee, Theo Horsley, Zifan Carl Guo, Tara Rezaei Kheirkhah, Qinyi Sun, Will Hathaway, Neel Nanda, Dimitris Bertsimas

专题命中 知识编辑与模型理解 :language model(title);分类 cs.CL、cs.AI、cs.LG

详情

展开后加载摘要…

URL PDF HTML 收藏
2302.09587 2024-01-09 cs.SE 78%

On the Reliability and Explainability of Language Models for Program Generation

Yue Liu, Chakkrit Tantithamthavorn, Yonghui Liu, Li Li

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments Accepted by ACM Transactions on Software Engineering and Methodology (TOSEM)

详情

展开后加载摘要…

URL PDF HTML 收藏
2311.15543 2023-11-28 cs.CV 78%

Beyond Pixels: Exploring Human-Readable SVG Generation for Simple Images with Vision Language Models

Tong Zhang, Haoyang Liu, Peiyan Zhang, Yuxuan Cheng, Haohan Wang

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments 10 pages, 7 figures

详情

展开后加载摘要…

URL PDF HTML 收藏
2303.12188 2023-08-03 cond-mat.mtrl-sci physics.comp-ph 78%

Toward Accurate Interpretable Predictions of Materials Properties within Transformer Language Models

Vadim Korolev, Pavel Protsenko

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments 17 pages, 5 figures, 1 table

详情

展开后加载摘要…

URL PDF HTML 收藏
2304.03307 2023-04-10 cs.CV eess.IV 78%

Vita-CLIP: Video and text adaptive CLIP via Multimodal Prompting

Syed Talal Wasim, Muzammal Naseer, Salman Khan, Fahad Shahbaz Khan, Mubarak Shah

专题命中 知识编辑与模型理解 :prompting(title,abstract)

Comments Accepted at CVPR-2023. Codes/models available at https://github.com/TalalWasim/Vita-CLIP

详情

展开后加载摘要…

URL PDF HTML 收藏
2301.00182 2023-03-28 cs.CV 78%

Bidirectional Cross-Modal Knowledge Exploration for Video Recognition with Pre-trained Vision-Language Models

Wenhao Wu, Xiaohan Wang, Haipeng Luo, Jingdong Wang, Yi Yang, Wanli Ouyang

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments Accepted by CVPR 2023

详情

展开后加载摘要…

URL PDF HTML 收藏
2301.11100 2023-01-27 cs.CV cs.CY cs.HC 78%

Vision-Language Models Performing Zero-Shot Tasks Exhibit Gender-based Disparities

Melissa Hall, Laura Gustafson, Aaron Adcock, Ishan Misra, Candace Ross

专题命中 知识编辑与模型理解 :language model(title,abstract)

详情

展开后加载摘要…

URL PDF HTML 收藏
2211.10578 2022-12-13 cs.CV 78%

ABINet++: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Spotting

Shancheng Fang, Zhendong Mao, Hongtao Xie, Yuxin Wang, Chenggang Yan, Yongdong Zhang

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments Accepted by TPAMI. Code is available at https://github.com/FangShancheng/ABINet-PP. arXiv admin note: substantial text overlap with arXiv:2103.06495 (conference version)

详情

展开后加载摘要…

URL PDF HTML 收藏
2005.13407 2022-11-15 cs.CL cs.AI cs.LG 78%

CausaLM: Causal Model Explanation Through Counterfactual Language Models

Amir Feder, Nadav Oved, Uri Shalit, Roi Reichart

专题命中 知识编辑与模型理解 :language model(title);分类 cs.CL、cs.AI、cs.LG

Comments Our code and data are available at: https://amirfeder.github.io/CausaLM/ Accepted for publication in Computational Linguistics journal

详情

展开后加载摘要…

URL PDF HTML 收藏
2203.14940 2022-03-29 cs.CV 78%

Learning to Prompt for Open-Vocabulary Object Detection with Vision-Language Model

Yu Du, Fangyun Wei, Zihe Zhang, Miaojing Shi, Yue Gao, Guoqi Li

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments Accepted by CVPR 2022

详情

展开后加载摘要…

URL PDF HTML 收藏
2103.06495 2021-03-12 cs.CV 78%

Read Like Humans: Autonomous, Bidirectional and Iterative Language Modeling for Scene Text Recognition

Shancheng Fang, Hongtao Xie, Yuxin Wang, Zhendong Mao, Yongdong Zhang

专题命中 知识编辑与模型理解 :language model(title,abstract)

Comments Accepted by CVPR 2021

详情

展开后加载摘要…

URL PDF HTML 收藏
1610.00735 2016-10-05 cs.IR 78%

MatLM: a Matrix Formulation for Probabilistic Language Models

Yanshan Wang, Hongfang Liu

专题命中 知识编辑与模型理解 :language model(title,abstract)

详情

展开后加载摘要…

URL PDF HTML 收藏
2508.17324 2025-10-02 cs.CL cs.AI 77%

CultranAI at PalmX 2025: Data Augmentation for Cultural Knowledge Representation

Hunzalah Hassan Bhatti, Youssef Ahmed, Md Arid Hasan, Firoj Alam

专题命中 知识编辑与模型理解 :large language model(abstract,comments);language model(abstract,comments);分类 cs.CL、cs.AI;foundation model(comments)

Comments LLMs, Native, Arabic LLMs, Augmentation, Multilingual, Language Diversity, Contextual Understanding, Minority Languages, Culturally Informed, Foundation Models, Large Language Models

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.15218 2026-08-24 cs.AI cs.CR 版本更新 77%

When Words Are Safe But Actions Kill: Probing Physical Jailbreak Beyond Textual Jailbreak in Hidden-State Risk Space

当言语安全但行动致命:在隐藏状态风险空间中探究超越文本安全的物理危险

Weimeng Wang, Ziqiang Wang, Zihang Zhan, Chuanpu Fu, Qi Li, Ke Xu

机构 * Tsinghua University(清华大学) Nanyang Technological University(南洋理工大学)

专题命中 知识编辑与模型理解 :LLM(abstract);large language model(abstract);language model(abstract);分类 cs.AI

AI总结 研究大语言模型中语言无害指令在现实世界中的物理危险与文本级内容危险是否相同,提出PRISM方法,通过隐藏状态方向分析等证明CD和PD可分离,PRISM在多个基准测试中表现良好,能检测物理危险而非仅依赖明确不安全措辞。

详情

展开后加载摘要…

URL PDF HTML 收藏
2608.13072 2026-08-14 cs.AI 新提交 77%

EEG-PRIME: Prototype-Aligned Representation Learning with Multi-Level Conditioning for EEG Decoding

EEG-PRIME:面向脑电信号解码的多层级条件原型对齐表示学习

Shuailei Zhang, Muyun Jiang, Wei Zhang, Jinbo Chen, Zhiwei Guo, Yong Li, Yi Ding, Cuntai Guan

机构 * College of Computing and Data Science, Nanyang Technological University(南洋理工大学计算与数据科学学院) Centre for AI in Medicine, Nanyang Technological University(南洋理工大学医学人工智能中心) Southeast University(东南大学)

专题命中 知识编辑与模型理解 :foundation model(abstract);instruction tuning(abstract);pretraining(abstract);分类 cs.AI

AI总结 本文提出EEG-PRIME脑电基础模型,结合掩码预训练与原型对齐指令调优,在16个多类型数据集上实现跨被试解码性能提升,且具备零样本迁移能力。

详情

展开后加载摘要…

URL PDF HTML 收藏
2608.09095 2026-08-11 cs.AI 新提交 77%

Who Bridges Safety? Identifying and Targeting Cross-Lingual Shared Safety Pathways

谁搭建安全桥梁?识别并靶向跨语言共享安全路径

Shuyi Miao, Wangjie Qiu, Pengyang Shao, Canran Xiao, Fei Shen, Zhiming Zheng, Tat-Seng Chua

专题命中 知识编辑与模型理解 :LLM(abstract_cn);large language model(abstract);language model(abstract);分类 cs.AI

AI总结 本研究识别出跨语言共享安全路径,提出基于该路径的靶向对齐方法,仅更新少量参数即可提升非高资源语言安全性并保留模型通用能力。

详情

展开后加载摘要…

URL PDF HTML 收藏
2608.04378 2026-08-06 cs.SD cs.LG eess.AS 新提交 77%

Helping Music Co-Creation Agents 'Listen' Well: Hierarchical Self-Supervised World Models for Understanding and Generation

助力音乐协同创作智能体“良好聆听”:用于理解与生成的分层自监督世界模型

Scott H. Hawley

机构 * Belmont University(贝尔蒙特大学)

专题命中 知识编辑与模型理解 :LLM(abstract,abstract_cn);prompting(abstract);分类 cs.LG

AI总结 本研究提出分层自监督世界模型,通过Swin V2编码器与条件流匹配模型构建协同音乐创作智能体,提升了和弦与调式检测准确率,可快速生成音乐建议并支持交互演示。

Comments 20 pages, 14 figures. A 6-page version was submitted to the NeurIPS 2026 Creative AI Track. Supplemental website with listening examples: https://drscotthawley.github.io/midi-rae-jepa-son/. Live demo: https://drscotthawley-midi-rae-jepa-son.hf.space/

详情

展开后加载摘要…

URL PDF HTML 收藏
2606.04552 2026-08-06 cs.CL q-bio.GN 版本更新 77%

LDARNet: DNA Adaptive Representation Network with Learnable Tokenization for Genomic Modeling

LDARNet: 用于基因组建模的DNA自适应表示网络与可学习分词

Daria Ledneva, Denis Kuznetsov

机构 * University of California, Berkeley(加州大学伯克利分校)

专题命中 知识编辑与模型理解 :large language model(abstract);language model(abstract);foundation model(abstract);分类 cs.CL

AI总结 提出LDARNet,一种结合动态分块和双向路由的120M参数层次基因组基础模型,在27个任务中优于更大模型,并发现学习到的边界与生物学基序对齐。

Comments Fix parameter count typo; update results

详情

展开后加载摘要…

URL PDF HTML 收藏
2608.00147 2026-08-04 cs.CV cs.LG 新提交 77%

RadPRISM: Schema-stratified radiology-report supervision for concept-disentangled image representations and visual grounding

RadPRISM:用于概念解耦图像表示与视觉定位的模式分层放射学报告监督方法

Fabian Drexel, Marlene Fritzsche, Era Stambollxhiu, Miriam Kumpf, Lena Schmitzer, Lea Schumann, Jannik Kahmann, Friedrich Puttkammer, Johannes Moll, Jannik Lübberstedt, Zeineb Ben Chaaben, Anirudh Narayanan, Cosmin I. Bercea, Sebastian Ziegelmayer, Marcus R. Makowski, Daniel Rueckert, Lisa C. Adams, Keno K. Bressem

机构 * Technical University of Munich (TUM)(慕尼黑工业大学(TUM)) TUM University Hospital(慕尼黑工业大学医院) Technical University of Munich, School of Medicine and Health(慕尼黑工业大学医学与健康学院) Klinikum rechts der Isar(右伊萨尔医院) Charité – Universitätsmedizin Berlin(柏林夏里特医学院) Freie Universität Berlin(柏林自由大学) Humboldt Universität zu Berlin(柏林洪堡大学) Imperial College London(伦敦帝国理工学院) Munich Center for Machine Learning (MCML)(慕尼黑机器学习中心(MCML)) University Hospital Essen (AöR)(埃森大学医院(AöR)) Institute for Artificial Intelligence in Medicine (IKIM)(医学人工智能研究所(IKIM)) Institute of Interventional and Diagnostic Radiology and Neuroradiology(介入与诊断放射学及神经放射学研究所) National Center for Tumor Diseases West(西部肿瘤疾病国家中心)

专题命中 知识编辑与模型理解 :large language model(abstract);language model(abstract);pretraining(abstract);分类 cs.LG

AI总结 RadPRISM将放射学模式作为分层轴,通过专用视觉子空间对齐临床概念,提升零样本分类与视觉定位性能,实现可透明检查的概念解耦医学图像表示。

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.01557 2026-07-31 cs.CL 版本更新 77%

Language Diversity: Evaluating Language Usage and AI Performance on African Languages in Digital Spaces

语言多样性:评估非洲语言在数字空间中的使用情况和人工智能性能

Edward Ajayi, Eudoxie Umwari, Mawuli Deku, Prosper Singadi, Jules Udahemuka, Bekalu Tadele, Chukuemeka Edeh

机构 * Carnegie Mellon University(卡内基梅隆大学) Africa Kigali Rwanda(非洲基加利卢旺达) Bahir Dar Institute of Technology(巴希尔达尔技术学院) Federal University Otuoke Bayelsa Nigeria(贝莱萨州联邦大学)

专题命中 知识编辑与模型理解 :LLM(abstract,abstract_cn);language model(abstract);分类 cs.CL

AI总结 本研究发现本地新闻数据比对话平台数据更有效用于训练非洲语言的AI模型,强调了处理干净和语言切换文本的重要性。

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.24750 2026-07-29 cs.CL cs.HC 新提交 77%

TimeCapsule: Generative Hallucination as a Method for Historical Sensemaking

时间胶囊:将生成性幻觉作为历史意义建构的一种方法

Hayk Grigorian, Hamed Yaghoobian

机构 * Muhlenberg College(Muhlenberg学院)

专题命中 知识编辑与模型理解 :large language model(abstract);language model(abstract);pretraining(abstract);分类 cs.CL

AI总结 研究针对大语言模型在叙述过去时不可靠的问题,提出在维多利亚时代文本上训练的TimeCapsule模型,通过定量评估和定性探究展示其性能及问题,认为对未来的无知使幻觉成为对19世纪本体论的解释性探索。

Comments 10 pages, 4 figures. Accepted to Creativity and Cognition (C&C '26)

详情

展开后加载摘要…

URL PDF HTML 收藏
2607.22572 2026-07-28 cs.AI 新提交 77%

Schema-Aware Localisation (SAL): Live Schema Grounding and Hallucination Validation for Oracle NL2SQL

模式感知本地化(SAL):用于Oracle NL2SQL的实时模式基础与幻觉验证

Sanjay Mishra, Divya Chukkapalli, Ganesh R. Naik

机构 * Torrens University Australia(澳大利亚托伦斯大学)

专题命中 知识编辑与模型理解 :LLM(abstract);large language model(abstract);language model(abstract);分类 cs.AI

AI总结 研究大型语言模型在Oracle数据库执行SQL失败的问题,提出模式感知本地化(SAL)轻量级中间件,通过构建实时模式映射、注入上下文及幻觉索引验证等方法,提升执行基础真值,减少执行失败率。

Comments 18 pages , 3 figures , 14 tables ,1 algorithm

详情

展开后加载摘要…

URL PDF HTML 收藏
2604.01457 2026-07-28 cs.CL 版本更新 77%

Wired for Overconfidence: A Mechanistic Perspective on Inflated Verbalized Confidence in LLMs

为自信所束缚:对大语言模型中过度自信的机械视角分析

Tianyi Zhao, Yinhan He, Wendy Zheng, Yujie Zhang, Chen Chen

机构 * University of Virginia(弗吉尼亚大学)

专题命中 知识编辑与模型理解 :LLM(abstract_cn);large language model(abstract);language model(abstract);分类 cs.CL

AI总结 研究探讨了大语言模型中过度自信的机制,发现中间层的MLP块和注意力头会生成自信信号,并通过干预提升校准性。

Comments COLM 2026

详情

展开后加载摘要…

URL PDF HTML 收藏