arXivDaily arXiv每日学术速递 周一至周五更新

期刊&会议

NeurIPS

Conference on Neural Information Processing Systems · 会议 · Machine Learning

2026-01-27 至 2026-01-27 共收录 16
2601.18733 2026-01-27 cs.RO cs.AI cs.CV

Advances and Innovations in the Multi-Agent Robotic System (MARS) Challenge

多智能体机器人系统(MARS)挑战的进展与创新

Li Kang, Heng Zhou, Xiufeng Song, Rui Li, Bruno N. Y. Chen, Ziye Wang, Ximeng Meng, Stone Tao, Yiran Qin, Xiaohong Liu, Ruimao Zhang, Lei Bai, Yilun Du, Hao Su, Philip Torr, Zhenfei Yin, Ruihao Gong, Yejun Zeng, Fengjun Zhong, Shenghao Jin, Jinyang Guo, Xianglong Liu, Xiaojun Jia, Tianqi Shan, Wenqi Ren, Simeng Qin, Jialing Yang, Xiaoyu Ma, Tianxing Chen, Zixuan Li, Zijian Cai, Yan Qin, Yusen Qin, Qiangyu Chen, Kaixuan Wang, Zhaoming Han, Yao Mu, Ping Luo, Yuanqi Yao, Haoming Song, Jan-Nico Zaech, Fabien Despinoy, Danda Pani Paudel, Luc Van Gool

机构 * SJTU(上海交通大学) Oxford(牛津大学) USTC(中国科学技术大学) Shanghai AI Lab(上海人工智能实验室) CMU(卡内基梅隆大学) HKU(香港大学) Tongji(同济大学) UC San Diego(南加州大学) CUHK-SZ(香港中文大学(深圳)) SYSU(华南理工大学) Harvard(哈佛大学)

AI总结 MARS挑战通过多智能体具身规划与控制任务,推动多智能体协作AI系统的发展。

Comments MARS Challenge @ NeurIPS 2025 Workshop on Space in Vision, Language, and Embodied AI. Challenge page: https://mars-eai.github.io/MARS-Challenge-Webpage/

详情

展开后加载摘要…

URL PDF HTML 收藏
2601.18076 2026-01-27 cs.LG

Comparison requires valid measurement: Rethinking attack success rate comparisons in AI red teaming

比较需要有效的测量:重新思考AI红队中的攻击成功率比较

Alexandra Chouldechova, A. Feder Cooper, Solon Barocas, Abhinav Palia, Dan Vann, Hanna Wallach

机构 * Microsoft Research(微软研究院)

AI总结 本文重新审视AI红队中攻击成功率的比较,指出其有效性问题并提出测量理论和统计学方法以评估比较的合理性。

Journal ref NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.20984 2026-01-27 cs.LG cs.AI

Learning Grouped Lattice Vector Quantizers for Low-Bit LLM Compression

学习分组晶格向量量化器以实现低比特LLM压缩

Xi Zhang, Xiaolin Wu, Jiamang Wang, Weisi Lin

机构 * Nanyang Technological University(南洋理工大学) Alibaba Group(阿里巴巴集团) Southwest Jiaotong University(西南交通大学)

AI总结 本文提出分组晶格向量量化器,通过学习生成矩阵优化低比特LLM压缩,实现模型大小与精度的平衡。

Comments NeurIPS 2025 Poster

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.14125 2026-01-27 cs.LG cs.AI q-bio.NC

Contrastive Consolidation of Top-Down Modulations Achieves Sparsely Supervised Continual Learning

对比整合上行调节实现稀疏监督连续学习

Viet Anh Khoa Tran, Emre Neftci, Willem A. M. Wybo

AI总结 任务调节对比学习通过上行调节实现稀疏监督下的连续学习,提升类别增量和迁移学习性能。

Comments Accepted to NeurIPS 2025. Camera-ready version. 33 pages, 5 figures. Updated acknowledgements. Code available at: https://github.com/tran-khoa/tmcl

详情

展开后加载摘要…

URL PDF HTML 收藏
2512.10046 2026-01-27 cs.AI

SimWorld-Robotics: Synthesizing Photorealistic and Dynamic Urban Environments for Multimodal Robot Navigation and Collaboration

SimWorld-Robotics: 为多模态机器人导航与协作合成逼真动态城市环境

Yan Zhuang, Jiawei Ren, Xiaokang Ye, Jianzhi Shen, Ruixuan Zhang, Tianai Yue, Muhammad Faayez, Xuhong He, Ziqiao Ma, Lianhui Qin, Zhiting Hu, Tianmin Shu

机构 * University of Virginia(弗吉尼亚大学) UC San Diego(加州大学圣地亚哥分校) Johns Hopkins University(约翰霍普金斯大学) Carnegie Mellon University(卡内基梅隆大学) University of Michigan(密歇根大学)

AI总结 SimWorld-Robotics通过合成逼真动态城市环境,提出两个多模态机器人基准测试,评估机器人在复杂场景中的导航、协作与通信能力。

Comments Conference: NeurIPS 2025 (main)

详情

展开后加载摘要…

URL PDF HTML 收藏
2511.10683 2026-01-27 cs.LG cs.CV

LT-Soups: Bridging Head and Tail Classes via Subsampled Model Soups

LT-Soups: 通过子采样模型汤桥接头部和尾部类别

Masih Aminbeidokhti, Subhankar Roy, Eric Granger, Elisa Ricci, Marco Pedersoli

机构 * École de technologie supérieure University of Bergamo(贝拉姆博大学) University of Trento(特伦托大学) Fondazione Bruno Kessler (FBK)(布鲁诺·凯斯勒基金会)

AI总结 LT-Soups通过两阶段模型汤框架在长尾分布中平衡头部和尾部类别性能,提升模型在不平衡数据集上的泛化能力。

Comments Neurips 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2510.22860 2026-01-27 cs.CL q-bio.NC

Far from the Shallow: Brain-Predictive Reasoning Embedding through Residual Disentanglement

远离浅层:通过残差解耦实现脑预测推理嵌入

Linyang He, Tianjun Zhong, Richard Antonello, Gavin Mischler, Micah Goldblum, Nima Mesgarani

机构 * Zuckerman Mind Brain Behavior Institute, Columbia University(扎克曼脑行为研究所,哥伦比亚大学) Department of Electrical Engineering, Columbia University(电气工程系,哥伦比亚大学) Department of Computer Science, Columbia University(计算机科学系,哥伦比亚大学)

AI总结 通过残差解耦方法,研究实现了对语言推理过程的神经嵌入解耦,揭示了推理在大脑中的独特预测能力及时间特征。

Comments Accepted at NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.21155 2026-01-27 cs.CL

Learning the Wrong Lessons: Syntactic-Domain Spurious Correlations in Language Models

学习错误的教训:语言模型中的语法-领域虚假相关性

Chantal Shaib, Vinith M. Suriyakumar, Levent Sagun, Byron C. Wallace, Marzyeh Ghassemi

机构 * Northeastern University(东北大学) MIT(麻省理工学院) Meta

AI总结 研究揭示了语言模型中语法与领域之间的虚假相关性问题,指出训练数据中语法模板可能影响模型性能,并提出需测试此类相关性及确保训练数据多样性以防止错误学习。

Comments NeurIPS 2025 Spotlight

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.17429 2026-01-27 cs.CV

Multi-scale Temporal Prediction via Incremental Generation and Multi-agent Collaboration

多尺度时间预测 via 逐步生成和多智能体协作

Zhitao Zeng, Guojian Yuan, Junyuan Mao, Yuxuan Wang, Xiaoshuang Jia, Yueming Jin

机构 * National University of Singapore(新加坡国立大学) Alibaba Group(阿里巴巴集团) Renmin University of China(中国人民大学)

AI总结 本文提出了一种多尺度时间预测方法,通过逐步生成和多智能体协作,提升多尺度和多状态预测的准确性和一致性。

Comments 20 pages, 6 figures

Journal ref NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2509.14275 2026-01-27 cs.CR cs.AI cs.CL cs.LG

FedMentor: Domain-Aware Differential Privacy for Heterogeneous Federated LLMs in Mental Health

FedMentor: 域感知差分隐私用于心理健康领域异构联邦大语言模型

Nobin Sarwar, Shubhashis Roy Dipta

机构 * University of Maryland, Baltimore County(马里兰大学巴尔的摩县分校)

AI总结 FedMentor通过整合LoRA和域感知差分隐私,实现对心理健康领域异构联邦大语言模型的隐私保护微调,提升安全性并保持模型效用。

Comments NeurIPS 2025 GenAI4Health Workshop

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.20671 2026-01-27 cs.CV

IPFormer: Visual 3D Panoptic Scene Completion with Context-Adaptive Instance Proposals

IPFormer: 基于上下文自适应实例提案的视觉3D全景场景补全

Markus Gross, Aya Fahmy, Danit Niwattananan, Dominik Muhle, Rui Song, Daniel Cremers, Henri Meeß

机构 * Fraunhofer Institute IVI(弗劳恩霍夫研究所IVI) Technical University of Munich(慕尼黑技术大学) Munich Center for Machine Learning(慕尼黑机器学习中心) University of California, Los Angeles(加州大学洛杉矶分校)

AI总结 IPFormer通过上下文自适应实例提案实现视觉3D全景场景补全,提升场景理解与泛化能力。

Journal ref Advances in Neural Information Processing Systems (NeurIPS) 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2506.18951 2026-01-27 cs.DB cs.AI

SWE-SQL: Illuminating LLM Pathways to Solve User SQL Issues in Real-World Applications

SWE-SQL:揭示LLM解决现实应用中用户SQL问题的路径

Jinyang Li, Xiaolong Li, Ge Qu, Per Jacobsson, Bowen Qin, Binyuan Hui, Shuzheng Si, Nan Huo, Xiaohan Xu, Yue Zhang, Ziwei Tang, Yuanshuai Li, Florensia Widjaja, Xintong Zhu, Feige Zhou, Yongfeng Huang, Yannis Papakonstantinou, Fatma Ozcan, Chenhao Ma, Reynold Cheng

机构 * HKU STAR Lab(香港大学STAR实验室) Google Cloud(谷歌云) CUHKSZ(香港中文大学) CUHK(香港中文大学) THU(清华大学) The BIRD Team(BIRD团队)

AI总结 SWE-SQL通过引入BIRD-CRITIC基准和Six-Gym训练环境,提升开源模型解决SQL问题的能力,实现对复杂SQL调试任务的突破。

Comments 29 pages, 10 figures, NeurIPS 2025 Main

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.22323 2026-01-27 cs.CL

Advancing Expert Specialization for Better MoE

推动专家专业化以提升MoE

Hongcan Guo, Haolang Lu, Guoshun Nan, Bolun Chu, Jialin Zhuang, Yuan Yang, Wenhao Che, Xinye Cao, Sicong Leng, Qimei Cui, Xudong Jiang

机构 * Beijing University of Posts and Telecommunications(北京邮电大学) Nanyang Technological University(南洋理工大学)

AI总结 本文提出正交性损失和方差损失,提升MoE模型专家专业化,显著提高性能并保持负载平衡。

Comments 33pages, 6figures(Accepted by Neurips 2025 Oral)

详情

展开后加载摘要…

URL PDF HTML 收藏
2505.20655 2026-01-27 cs.CV

Photography Perspective Composition: Towards Aesthetic Perspective Recommendation

摄影视角构图:迈向审美视角推荐

Lujian Yao, Siming Zheng, Xinbin Yuan, Zhuoxuan Cai, Pu Wu, Jinwei Chen, Bo Li, Peng-Tao Jiang

机构 * vivo Mobile Communication Co., Ltd(vivo移动通信有限公司)

AI总结 本文提出摄影视角构图(PPC)方法,通过自动化数据集构建、视频生成和视角质量评估模型,解决视角变换数据稀缺和评估标准模糊的问题,帮助普通用户提升摄影构图能力。

Comments Accepted at NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2503.08305 2026-01-27 cs.LG physics.chem-ph physics.comp-ph

ELECTRA: A Cartesian Network for 3D Charge Density Prediction with Floating Orbitals

ELECTRA:一种用于3D电荷密度预测的笛卡尔网络

Jonas Elsborg, Luca Thiede, Alán Aspuru-Guzik, Tejs Vegge, Arghya Bhowmik

机构 * Technical University of Denmark(丹麦技术大学) CAPeX Pioneer Center for Accelerating P2X Materials Discovery(CAPeX先锋中心) University of Toronto(多伦多大学) Vector Institute for Artificial Intelligence(人工智能矢量研究所) Canadian Institute for Advanced Research (CIFAR)(加拿大高级研究 institute)

AI总结 ELECTRA通过笛卡尔张量网络预测漂浮轨道位置,实现高效且准确的3D电荷密度预测,并显著降低DFT计算的SCF迭代次数。

Comments 10 pages, 4 figures, 5 tables, NeurIPS 2025

详情

展开后加载摘要…

URL PDF HTML 收藏
2411.18624 2026-01-27 cs.CV

GeneMAN: Generalizable Single-Image 3D Human Reconstruction from Multi-Source Human Data

GeneMAN: 从多源人类数据中通用的单图像3D人体重建

Wentao Wang, Hang Ye, Fangzhou Hong, Xue Yang, Jianfu Zhang, Yizhou Wang, Ziwei Liu, Liang Pan

机构 * Shanghai AI Laboratory(上海人工智能实验室) Peking University(北京大学) Nanyang Technological University(南洋理工大学) SAIS & SCS, Shanghai Jiao Tong University(上海交通大学SAIS与SCS)

AI总结 GeneMAN通过多源高质量数据集和深度学习方法,实现从单张图像到高保真3D人体的通用重建。

Comments Accepted by NeurIPS 2025; Project page: https://roooooz.github.io/GeneMAN/

详情

展开后加载摘要…

URL PDF HTML 收藏