大模型对齐与安全

Towards Fair Graph Federated Learning via Incentive Mechanisms

Chenglu Pan, Jiarong Xu, Yue Yu, Ziqi Yang, Qingbiao Wu, Chunping Wang, Lei Chen, Yang Yang

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI、cs.LG

Comments 15 pages, 4 figures, accepted by AAAI'24

Progression and Challenges of IoT in Healthcare: A Short Review

S M Atikur Rahman, Sifat Ibtisum, Priya Podder, S. M. Saokat Hossain

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI、cs.LG

Comments 7 pages

Journal ref International Journal of Computer Applications, Vol. 185, No. 37, pp. 9-15, October 2023

SoK: Memorization in General-Purpose Large Language Models

Valentin Hartmann, Anshuman Suri, Vincent Bindschaedler, David Evans, Shruti Tople, Robert West

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.CL、cs.LG

Provable Robust Watermarking for AI-Generated Text

Xuandong Zhao, Prabhanjan Ananth, Lei Li, Yu-Xiang Wang

专题命中 隐私与版权 ：safety(abstract)；分类 cs.CL、cs.LG

Instruction-Following Speech Recognition

Cheng-I Jeff Lai, Zhiyun Lu, Liangliang Cao, Ruoming Pang

专题命中 隐私与版权 ：safety(abstract)；分类 cs.CL、cs.LG

Membership Inference Attacks against Language Models via Neighbourhood Comparison

Justus Mattern, Fatemehsadat Mireshghallah, Zhijing Jin, Bernhard Schölkopf, Mrinmaya Sachan, Taylor Berg-Kirkpatrick

专题命中 隐私与版权 ：safety(abstract)；分类 cs.CL、cs.LG

Requirements for Explainability and Acceptance of Artificial Intelligence in Collaborative Work

Sabine Theis, Sophie Jentzsch, Fotini Deligiannaki, Charles Berro, Arne Peter Raulf, Carmen Bruder

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI、cs.CY

Personalized DP-SGD using Sampling Mechanisms

Geon Heo, Junseok Seo, Steven Euijong Whang

专题命中 隐私与版权 ：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments 10 pages, 5 figures

FaRO 2: an Open Source, Configurable Smart City Framework for Real-Time Distributed Vision and Biometric Systems

Joel Brogan, Nell Barber, David Cornett, David Bolme

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI、cs.LG

Nebula-I: A General Framework for Collaboratively Training Deep Learning Models on Low-Bandwidth Cloud Clusters

Yang Xiang, Zhihua Wu, Weibao Gong, Siyu Ding, Xianjie Mo, Yuang Liu, Shuohuan Wang, Peng Liu, Yongshuai Hou, Long Li, Bin Wang, Shaohuai Shi, Yaqian Han, Yue Yu, Ge Li, Yu Sun, Yanjun Ma, Dianhai Yu

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI、cs.LG

Comments 20 pages, 10 figures, technical report

Leveraging Privacy Profiles to Empower Users in the Digital Society

Davide Di Ruscio, Paola Inverardi, Patrizio Migliarini, Phuong T. Nguyen

专题命中 隐私与版权 ：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments The paper consists of 37 pages, 11 figures

Towards Personalized Federated Learning

Alysa Ziying Tan, Han Yu, Lizhen Cui, Qiang Yang

专题命中 隐私与版权 ：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments Accepted by IEEE Transactions on Neural Networks and Learning Systems

Towards a Secure and Reliable Federated Learning using Blockchain

Hajar Moudoud, Soumaya Cherkaoui, Lyes Khoukhi

专题命中 隐私与版权 ：trustworthy(abstract)；分类 cs.AI、cs.LG

Comments This paper has been accepted for publication by IEEE Global Communications Conference (GLOBECOM). The final version will be published by the IEEE

Multimodal Co-learning: Challenges, Applications with Datasets, Recent Advances and Future Directions

Anil Rahate, Rahee Walambe, Sheela Ramanna, Ketan Kotecha

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI、cs.LG

Comments This is published in Information Fusion Journal and published copy can be downloaded from https://authors.elsevier.com/c/1eIVS5a7-Gls0Y for 50 days. https://www.sciencedirect.com/science/article/pii/S1566253521002530

Journal ref Information Fusion 81(2022) 203-239

NeuCrowd: Neural Sampling Network for Representation Learning with Crowdsourced Labels

Yang Hao, Wenbiao Ding, Zitao Liu

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI、cs.LG

Comments Accepted in Knowledge and Information Systems

Knowledge-intensive Language Understanding for Explainable AI

Amit Sheth, Manas Gaur, Kaushik Roy, Keyur Faldu

专题命中 隐私与版权 ：trustworthy(abstract)；分类 cs.CL、cs.AI

Comments To appear in IEEE Internet Computing, September/October 2021 Issue

On the Privacy Risks of Algorithmic Fairness

Hongyan Chang, Reza Shokri

专题命中 隐私与版权 ：trustworthy(abstract)；分类 cs.CY、cs.LG

A Scalable Two Stage Approach to Computing Optimal Decision Sets

Alexey Ignatiev, Edward Lam, Peter J. Stuckey, Joao Marques-Silva

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI、cs.LG

Neural Machine Translation in Linear Time

Nal Kalchbrenner, Lasse Espeholt, Karen Simonyan, Aaron van den Oord, Alex Graves, Koray Kavukcuoglu

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.CL、cs.LG

Comments 9 pages

AlignFace: Human-Aligned Face Similarity Metric with Interpretable Concept Relations

AlignFace：具有可解释概念关系的人类对齐人脸相似度度量

Ying Huang, Wencan Zhang, Brian Y. Lim

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI

AI总结 本研究针对人脸生成模型评估中现有度量未实现认知对齐的问题，提出AlignFace度量，结合VLM、CA、CBM、GAM等技术，利用FACETS数据集，提升了与人类亚群感知的对齐度。

Comments 10 pages, 10 figures, 2 tables, ACM MM 26

Heterogeneity-Aware Belief Synchronization for Semantic Communication in AI-Native 6G Networks

面向AI原生6G网络中语义通信的异质性感知信念同步

Muhammad Hannan Akram, Muhammad Abubakar Rashid, Wassi Haider Kabir, Haejoon Jung, Kapal Dev, Syed Ali Hassan

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI

AI总结 针对AI原生6G网络中异构AI模型导致的信念对齐难题，提出基于MEC服务器潜在翻译模型的异质性感知信念同步框架，可降低同步成本、保护隐私并维持低信念对齐误差。

Comments 8 pages, 4 figures

MarkNull: Model-Agnostic Watermark Removal in AI-Generated Images via On-Manifold Latent Manipulation

MarkNull：通过流形上潜在空间操纵实现AI生成图像的模型无关水印去除

Jie Cao, Qi Li, Zelin Zhang, Xiaodong Wu, Lingshuang Liu, Xiangman Li, Jianbing Ni

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI

AI总结 本文提出MarkNull及变体MarkNull-A，通过流形上潜在空间操纵实现AI生成图像的模型无关水印去除，可降低水印比特准确率至近随机猜测水平，还能破解谷歌SynthID-Image系统，为防御提供攻击检测机制。

Comments Accepted to the 35th USENIX Security Symposium (USENIX Security 2026)

Software Engineering for and with GUI Agent

面向GUI智能体的软件工程及与GUI智能体协同的软件工程

Shengcheng Yu, Yuchen Ling, Junyang Xing, Quan Zhou, Chunrong Fang, Zhenyu Chen

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI

AI总结 该研究通过分析2018年1月至2026年4月的336篇GUI智能体论文，指出其在恢复机制、安全执行等方面的不足，提出需结合可靠执行与生命周期测试等以构建可部署的GUI智能体系统。

Privacy-Preserving Data Drift Detection and Recovery for Large-Scale LLM Applications via Proxy Representations

基于代理表示的大规模大语言模型应用的隐私保护数据漂移检测与恢复

Michael Levit, Josh Ledgard, Haoyu Dong, Vishwas Suryanarayanan, Eyal Kolman, Sharon Tan, Qiang Gan, Vishal Chowdhary

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI

AI总结 针对大规模LLM应用的隐私约束导致的评估与漂移追踪难题，提出ProxyDrift框架，基于非PII代理表示实现无敏感数据暴露的漂移监测与合成数据生成，实验验证其对齐度达RA~0.9。

EFFEKT: Efficient Federated Knowledge Transfer to Foundation Models

EFFEKT：面向基础模型的高效联邦知识迁移

Matteo Caligiuri, Francesco Barbato, Pietro Zanuttigh, Francesco Restuccia

机构 * Northeastern University（东北大学） ； University of Padua（帕多瓦大学） 纠错

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.LG

AI总结 EFFEKT是一种联邦学习框架，通过双向跨蒸馏策略，结合轻量级客户端代理模型与服务器端基础模型，实现高效的领域特定LoRA适配器训练，在低功耗边缘设备上性能优于基线。

Comments 12 main content pages, 8 appendix pages; 3 main figures, 9 appendix figures; 8 main tables, 9 appendix tables; 1 main algorithm, 4 appendix algorithms; accepted at TMLR

Copyright Is the Headline; Capability Is the Blind Spot: AI Technology in the Book-Publishing Trade Press, November 2025--August 2026

版权是头条，能力是盲区：2025年11月—2026年8月图书出版行业媒体中的人工智能技术

Fred Zimmerman

专题命中 隐私与版权 ：prompt injection(abstract)；分类 cs.CY

AI总结 本研究通过分析多国行业媒体的89篇文章，指出图书出版领域AI报道存在能力盲区，提出设立常设AI报道板块等改进建议。

Comments 7 pages, 2 figures. 15-page Supplemental Information, coded 89-item corpus, crosstabs, BibTeX database, and citation audit included as ancillary files

Chatting with Confidants or Corporations? Privacy Management with AI Companions

与知己还是公司聊天？AI伴侣的隐私管理

Hsuen-Chi Chiu, Jeremy Foote

专题命中 隐私与版权 ：safety(abstract)；分类 cs.CY

AI总结 基于沟通隐私管理理论和水平-垂直隐私框架，通过访谈Replika和Character.AI用户，发现用户融合人际习惯与机构意识，采用分层策略管理隐私，但拟人化设计导致边界模糊和隐私动荡。

Journal ref Journal of Computer-Mediated Communication, 2026

Edge Intelligence in Civil Aviation: Paradigms, Techniques, and Applications

民航中的边缘智能：范式、技术与应用

Wenbin Li, Zhongtian Liao, Bolin Liu, Yongjie Zhou, Jingling Wu, Xiaoyong Lin, Jing Chen

专题命中 隐私与版权 ：safety(abstract)；分类 cs.AI

AI总结 针对民航网络边缘数据处理问题，介绍边缘人工智能通过压缩、协作推理和分割学习等技术，减少延迟、带宽和暴露风险，实现断连平稳运行，阐述其运行动机、技术、范式及应用趋势，为民航提供低延迟、隐私保护和弹性的人工智能服务。

Comments 8 pages, 2 figures, 2 tables

Journal ref 2025 5th International Conference on Information Communication and Software Engineering (ICICSE), Chongqing, China, 12 to 14 December 2025, IEEE, 2026

Autoresearch with Coding Agents: Generalizers and Metric-Maximizers on Quran Recitation Data

使用编码智能体进行自动研究：古兰经诵读数据上的泛化器和指标最大化器

Nursultan Askarbekuly, Mohamad Al Mdfaa, Ahmed Helaly, Gonzalo Ferrer, Manuel Mazzara

机构 * Innopolis University（因诺波利斯大学） ； Skolkovo Institute of Science and Technology（斯克尔科沃科学与技术研究所） 纠错

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.AI

AI总结 研究在古兰经诵读数据任务中智能体自动研究模式，Claude Code和OpenAI Codex从同一起点出发，先发明相同算法后有分歧，添加测试集后情况变化，还提炼出评估自主智能体的五条设计规则，智能体解决方案超越手工管道。

In-Context Probing for Membership Inference in Fine-Tuned Language Models

上下文探查用于微调语言模型中的成员推断

Zhexi Lu, Hongliang Chi, Nathalie Baracaldo, Swanand Ravindra Kadhe, Yuseok Jeon, Lei Yu

机构 * Korea University, Seoul, South Korea（韩国大学） 纠错

专题命中 隐私与版权 ：alignment(abstract)；分类 cs.LG

AI总结 本文提出ICP-MIA框架，通过上下文探查技术有效检测微调语言模型中的成员身份，提升隐私保护能力。

Journal ref Proceedings of the 33rd Network and Distributed System Security Symposium (NDSS 2026)