NeurIPS

ROOT: Rethinking Offline Optimization as Distributional Translation via Probabilistic Bridge

Manh Cuong Dao, The Hung Tran, Phi Le Nguyen, Thao Nguyen Truong, Trong Nghia Hoang

机构 * National University of Singapore（新加坡国立大学） ； Washington State University（华盛顿州立大学） ； Hanoi University of Science and Technology（河内科学技术大学） ； National Institute of Advanced Industrial Science and Technology（国家先进工业科学与技术研究院） 纠错

Comments The first two authors contributed equally

Journal ref NeurIPS 2025 Spotlight

ViSpec: Accelerating Vision-Language Models with Vision-Aware Speculative Decoding

Jialiang Kang, Han Shu, Wenshuo Li, Yingjie Zhai, Xinghao Chen

机构 * Peking University（北京大学） ； Huawei Noah’s Ark Lab（华为诺亚实验室） 纠错

Comments NeurIPS 2025

Multi-Agent Reinforcement Learning for Task Offloading in Wireless Edge Networks

Andrea Fox, Francesco De Pellegrini, Eitan Altman

机构 * LIA, Avignon University（阿维尼翁大学LIA中心） ； INRIA, Sophia Antipolis, France（法国索菲亚Antipolis研究所） 纠错

Comments Oral presentation at AI4NextG @ NeurIPS'25 Workshop

Debate or Vote: Which Yields Better Decisions in Multi-Agent Large Language Models?

Hyeong Kyu Choi, Xiaojin Zhu, Sharon Li

机构 * Department of Computer Sciences, University of Wisconsin-Madison（计算机科学系，威斯康星大学麦迪逊分校） 纠错

Comments NeurIPS 2025 Spotlight

Optimal Dynamic Regret by Transformers for Non-Stationary Reinforcement Learning

Baiyuan Chen, Shinji Ito, Masaaki Imaizumi

机构 * The University of Tokyo（东京大学） ； RIKEN AIP（日本科学技术厅AIP） 纠错

Comments 27 pages

Journal ref NeurIPS 2025

Distributional Adversarial Attacks and Training in Deep Hedging

Guangyi He, Tobias Sutter, Lukas Gonon

机构 * Department of Mathematics Imperial College London（数学系帝国理工学院伦敦分校） ； Department of Economics University of St. Gallen（经济学系圣加尔登大学） ； School of Computer Science University of St. Gallen（计算机科学学院圣加尔登大学） 纠错

Comments Camera-ready version (accepted at NeurIPS 2025 https://neurips.cc/virtual/2025/poster/115434)

Learning Modular Exponentiation with Transformers

David Demitri Africa, Sara M. Kapoor, Theo Simon Sorg, Challenger Mishra

机构 * Department of Computer Science and Technology（计算机科学与技术系） 纠错

Comments Accepted at the 5th MATH-AI Workshop, NeurIPS'25

Flow based approach for Dynamic Temporal Causal models with non-Gaussian or Heteroscedastic Noises

Abdellah Rahmani, Pascal Frossard

机构 * LTS4, EPFL（EPFL 瑞士洛桑实验室） 纠错

Journal ref 39th Conference on Neural Information Processing Systems (NeurIPS 2025)

Watermarking Autoregressive Image Generation

Nikola Jovanović, Ismail Labiad, Tomáš Souček, Martin Vechev, Pierre Fernandez

机构 * Meta FAIR ； ETH Zurich 纠错

Comments NeurIPS 2025

Execution Guided Line-by-Line Code Generation

Boaz Lavon, Shahar Katz, Lior Wolf

机构 * Blavatnik School of Computer Science and AI, Tel Aviv University（巴尔-以色列计算机科学与人工智能学院，特拉维夫大学） 纠错

Comments Accepted to NeurIPS 2026

LeVo: High-Quality Song Generation with Multi-Preference Alignment

Shun Lei, Yaoxun Xu, Zhiwei Lin, Huaicheng Zhang, Wei Tan, Hangting Chen, Jianwei Yu, Yixuan Zhang, Chenyu Yang, Haina Zhu, Shuai Wang, Zhiyong Wu, Dong Yu

机构 * Shenzhen International Graduate School, Tsinghua University, Shenzhen（清华大学深圳国际研究生院） ； Tencent AI Lab（腾讯AI实验室） ； Wuhan University（武汉大学） ； The Chinese University of Hong Kong, Shenzhen (CUHK-Shenzhen)（香港中文大学（深圳）） ； X-LANCE Lab, Shanghai Jiao Tong University, Shanghai（上海交通大学X-LANCE实验室） ； School of Intelligence Science and Technology, Nanjing University, Suzhou, China（南京大学智能科学与技术学院） 纠错

Comments Accepted by NeurIPS 2025

Spark Transformer: Reactivating Sparsity in FFN and Attention

Chong You, Kan Wu, Zhipeng Jia, Lin Chen, Srinadh Bhojanapalli, Jiaxian Guo, Utku Evci, Jan Wassenberg, Praneeth Netrapalli, Jeremiah J. Willcock, Suvinay Subramanian, Felix Chern, Alek Andreev, Shreya Pathak, Felix Yu, Prateek Jain, David E. Culler, Henry M. Levy, Sanjiv Kumar

机构 * Google（谷歌） 纠错

Comments NeurIPS 2025

Does Thinking More always Help? Mirage of Test-Time Scaling in Reasoning Models

Soumya Suvra Ghosal, Souradip Chakraborty, Avinash Reddy, Yifu Lu, Mengdi Wang, Dinesh Manocha, Furong Huang, Mohammad Ghavamzadeh, Amrit Singh Bedi

机构 * University of Maryland（马里兰大学） ； Princeton University（普林斯顿大学） ； Capital One（Capital One公司） ； Amazon AGI ； University of Central Florida（佛罗里达中央大学） 纠错

Comments Accepted at NeurIPS 2025

Revisiting End-to-End Learning with Slide-level Supervision in Computational Pathology

Wenhao Tang, Rong Qin, Heng Fang, Fengtao Zhou, Hao Chen, Xiang Li, Ming-Ming Cheng

机构 * Nankai International Advanced Research Institute (Shenzhen Futian)（南开国际先进研究院（深圳福田）） ； VCIP, School of Computer Science, Nankai University（VCIP，计算机科学学院，南开大学） ； Huazhong University of Science and Technology（华中科技大学） ； The Hong Kong University of Science and Technology（香港科技大学） 纠错

Comments published on NeurIPS 2025

Blending Complementary Memory Systems in Hybrid Quadratic-Linear Transformers

Kazuki Irie, Morris Yau, Samuel J. Gershman

机构 * Harvard University（哈佛大学） ； MIT CSAIL（麻省理工学院计算机科学与人工智能实验室） ； Kempner Institute for the Study of Natural and Artificial Intelligence（自然与人工智能研究学院） 纠错

Comments Accepted to NeurIPS 2025

Lessons Learned: A Multi-Agent Framework for Code LLMs to Learn and Improve

Yuanzhe Liu, Ryan Deng, Tim Kaler, Xuhao Chen, Charles E. Leiserson, Yao Ma, Jie Chen

机构 * Rensselaer Polytechnic Institute（拉特兰理工学院） ； Massachusetts Institute of Technology（麻省理工学院） ； Michigan State University（密歇根州立大学） ； MIT-IBM Watson AI Lab, IBM Research（MIT-IBM沃森人工智能实验室，IBM研究院） 纠错

Comments NeurIPS 2025. Code is available at https://github.com/MITIBM-FastCoder/LessonL

BioCLIP 2: Emergent Properties from Scaling Hierarchical Contrastive Learning

Jianyang Gu, Samuel Stevens, Elizabeth G Campolongo, Matthew J Thompson, Net Zhang, Jiaman Wu, Andrei Kopanev, Zheda Mai, Alexander E. White, James Balhoff, Wasila Dahdul, Daniel Rubenstein, Hilmar Lapp, Tanya Berger-Wolf, Wei-Lun Chao, Yu Su

机构 * The Ohio State University（俄亥俄州立大学） ； Smithsonian Institution（史密森学会） ； UNC Chapel Hill（北卡罗来纳大学教堂山分校） ； University of California, Irvine（加州大学伊市分校） ； Princeton University（普林斯顿大学） ； Duke University（杜克大学） 纠错

Comments NeurIPS 2025 Spotlight; Project page: https://imageomics.github.io/bioclip-2/

REOrdering Patches Improves Vision Models

Declan Kutscher, David M. Chan, Yutong Bai, Trevor Darrell, Ritwik Gupta

机构 * University of Pittsburgh（匹兹堡大学） ； University of California, Berkeley（加州大学伯克利分校） 纠错

Comments Accepted to the 39th Conference on Neural Information Processing Systems (NeurIPS 2025)

PreFM: Online Audio-Visual Event Parsing via Predictive Future Modeling

Xiao Yu, Yan Fang, Xiaojie Jin, Yao Zhao, Yunchao Wei

机构 * Institute of Information Science, Beijing Jiaotong University（信息科学学院，北京交通大学） 纠错

Comments This paper is accepted by 39th Conference on Neural Information Processing Systems (NeurIPS 2025)

Sherlock: Self-Correcting Reasoning in Vision-Language Models

Yi Ding, Ruqi Zhang

机构 * Department of Computer Science, Purdue University, USA（计算机科学系，普渡大学） 纠错

Comments Published at NeurIPS 2025, 27 pages

Train with Perturbation, Infer after Merging: A Two-Stage Framework for Continual Learning

Haomiao Qiu, Miao Zhang, Ziyue Qiao, Liqiang Nie

机构 * Harbin Institute of Technology (Shenzhen)（哈尔滨工业大学（深圳）） ； Great Bay University（大湾大学） 纠错

Comments Accepted by NeurIPS 2025

Balanced Token Pruning: Accelerating Vision Language Models Beyond Local Optimization

Kaiyuan Li, Xiaoyue Chen, Chen Gao, Yong Li, Xinlei Chen

机构 * Tsinghua Shenzhen International Graduate School（清华大学深圳国际研究生院） ； BNRist, Tsinghua University（清华大学北京研究院） 纠错

Comments Accepted by Neurips 2025

Born a Transformer -- Always a Transformer? On the Effect of Pretraining on Architectural Abilities

Mayank Jobanputra, Yana Veitsman, Yash Sarrof, Aleksandra Bakalova, Vera Demberg, Ellie Pavlick, Michael Hahn

机构 * Saarland University（萨尔兰大学） ； Brown University（布朗大学） 纠错

Comments NeurIPS 2025

PolyPose: Deformable 2D/3D Registration via Polyrigid Transformations

Vivek Gopalakrishnan, Neel Dey, Polina Golland

机构 * MIT（麻省理工学院） 纠错

Comments NeurIPS 2025. Code available at https://github.com/eigenvivek/polypose

On the Emergence of Linear Analogies in Word Embeddings

Daniel J. Korchinski, Dhruva Karkada, Yasaman Bahri, Matthieu Wyart

机构 * Google DeepMind（谷歌DeepMind） ； Johns Hopkins & EPFL（约翰霍普金斯大学与EPFL） 纠错

Comments Main: 10 pages, 3 figures. Appendices: 11 pages, 7 figures. Accepted at NeurIPS 2025 as a poster

Hybrid Latent Reasoning via Reinforcement Learning

Zhenrui Yue, Bowen Jin, Huimin Zeng, Honglei Zhuang, Zhen Qin, Jinsung Yoon, Lanyu Shang, Jiawei Han, Dong Wang

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； Google（谷歌） ； LMU（莱比锡马克斯·普朗克研究所） 纠错

Comments NeurIPS 2025

DesignX: Human-Competitive Algorithm Designer for Black-Box Optimization

Hongshu Guo, Zeyuan Ma, Yining Ma, Xinglin Zhang, Wei-Neng Chen, Yue-Jiao Gong

机构 * South China University of Technology（南方科技大学） ； Massachusetts Institute of Technology（麻省理工学院） 纠错

Comments Accepted by NeurIPS 2025

MODEM: A Morton-Order Degradation Estimation Mechanism for Adverse Weather Image Recovery

Hainuo Wang, Qiming Hu, Xiaojie Guo

机构 * College of Intelligence and Computing, Tianjin University（智能与计算学院，天津大学） 纠错

Comments Accepted by NeurIPS 2025

Wasserstein Transfer Learning

Kaicheng Zhang, Sinian Zhang, Doudou Zhou, Yidong Zhou

机构 * School of Mathematical Sciences, Zhejiang University, China（浙江大学数学科学学院） ； Division of Biostatistics and Health Data Science, University of Minnesota, USA（明尼苏达大学生物统计与健康数据科学系） ； Department of Statistics and Data Science, National University of Singapore, Singapore（新加坡国立大学统计与数据科学系） ； Department of Statistics, University of California, Davis, USA（加州大学戴维斯分校统计系） 纠错

Comments 25 pages, 6 figures, NeurIPS 2025

Tropical Attention: Neural Algorithmic Reasoning for Combinatorial Algorithms

Baran Hashemi, Kurt Pasque, Chris Teska, Ruriko Yoshida

机构 * Origins Data Science Lab（Origins数据科学实验室） ； Technical University of Munich（慕尼黑技术大学） ； Naval Postgraduate School（海军研究生学院） 纠错

Comments Published at NeurIPS 2025