NeurIPS

Large Language Models as Medical Codes Selectors: a benchmark using the International Classification of Primary Care

Vinicius Anjos de Almeida, Vinicius de Camargo, Raquel Gómez-Bravo, Egbert van der Haring, Kees van Boven, Marcelo Finger, Luis Fernandez Lopez

机构 * Medical School, University of São Paulo（圣保罗大学医学院） ； University of São Paulo（圣保罗大学） ； Department of Epidemiology, School of Public Health, University of São Paulo（圣保罗大学流行病学系） ； Rehaklinik, Centre Hospitalier Neuro-psychiatrique (CHNP)（康复诊所，神经精神病中心（CHNP）） ； Radboud University（拉德堡德大学） ； Department of Primary and Community Care, Radboud University（初级与社区护理系，拉德堡德大学） ； Institute of Mathematics and Statistics, University of Sao Paulo（数学与统计学研究所，圣保罗大学） 纠错

Comments Accepted at NeurIPS 2025 as a poster presentation in The Second Workshop on GenAI for Health: Potential, Trust, and Policy Compliance (https://openreview.net/forum?id=Kl7KZwJFEG). 33 pages, 10 figures (including appendix), 15 tables (including appendix). To be submitted to peer-reviewed journal. For associated code repository, see https://github.com/almeidava93/llm-as-code-selectors-paper

MOSPA: Human Motion Generation Driven by Spatial Audio

Shuyang Xu, Zhiyang Dou, Mingyi Shi, Liang Pan, Leo Ho, Jingbo Wang, Yuan Liu, Cheng Lin, Yuexin Ma, Wenping Wang, Taku Komura

机构 * The University of Hong Kong（香港大学） ； Shanghai AI Lab（上海人工智能实验室） ； The Hong Kong University of Science and Technology（香港科学与技术大学） ； Macau University of Science and Technology（澳门科学理工大学） ； ShanghaiTech University（上海科技大学） ； Texas A&M University（德克萨斯A&M大学） 纠错

Comments NeurIPS 2025 (Spotlight)

Through the River: Understanding the Benefit of Schedule-Free Methods for Language Model Training

Minhak Song, Beomhan Baek, Kwangjun Ahn, Chulhee Yun

机构 * KAIST（韩国科学技术院） ； SNU & KAIST InnoCORE LLM（首尔国立大学及韩国科学技术院InnoCORE LLM） ； Microsoft Research（微软研究院） 纠错

Comments Published at NeurIPS 2025

Schrödinger Bridge Matching for Tree-Structured Costs and Entropic Wasserstein Barycentres

Samuel Howard, Peter Potaptchik, George Deligiannidis

机构 * Department of Statistics University of Oxford（统计系牛津大学） 纠错

Comments NeurIPS 2025

Over-squashing in Spatiotemporal Graph Neural Networks

Ivan Marisca, Jacob Bamberger, Cesare Alippi, Michael M. Bronstein

机构 * Università della Svizzera italiana, IDSIA, Lugano, Switzerland（瑞士大学、IDSIA、卢加诺，瑞士） ； University of Oxford, Oxford, UK（牛津大学、牛津，英国） ； Politecnico di Milano, Milan, Italy（米兰理工大学、米兰，意大利） ； AITHYRA, Vienna, Austria（AITHYRA、维也纳，奥地利） 纠错

Comments Accepted at NeurIPS 2025

Anti-Aliased 2D Gaussian Splatting

Mae Younes, Adnane Boukhayma

机构 * INRIA France（法国INRIA） ； University of Rennes（里昂大学） ； CNRS（国家科学研究中心） ； IRISA（里莫斯岛研究所） 纠错

Comments NeurIPS 2025. Code will be available at https://github.com/maeyounes/AA-2DGS

Where and How to Perturb: On the Design of Perturbation Guidance in Diffusion and Flow Models

Donghoon Ahn, Jiwon Kang, Sanghyun Lee, Minjae Kim, Jaewon Min, Wooseok Jang, Sangwu Lee, Sayak Paul, Susung Hong, Seungryong Kim

Comments Accepted at NeurIPS 2025. Project page: https://cvlab-kaist.github.io/HeadHunter/

SPARTA ALIGNMENT: Collectively Aligning Multiple Language Models through Combat

Yuru Jiang, Wenxuan Ding, Shangbin Feng, Greg Durrett, Yulia Tsvetkov

机构 * Zhejiang University（浙江大学） ； New York University（纽约大学） ； University of Washington（华盛顿大学） 纠错

Comments NeurIPS 2025

Large Stepsizes Accelerate Gradient Descent for Regularized Logistic Regression

Jingfeng Wu, Pierre Marion, Peter Bartlett

机构 * UC Berkeley（加州大学伯克利分校） ； Inria（法国国家科学与技术研究院） ； PSL University（巴黎社会科学高等学院） ； Google DeepMind（谷歌DeepMind） 纠错

Comments NeurIPS 2025 camera ready version

TiRex: Zero-Shot Forecasting Across Long and Short Horizons with Enhanced In-Context Learning

Andreas Auer, Patrick Podest, Daniel Klotz, Sebastian Böck, Günter Klambauer, Sepp Hochreiter

机构 * NXAI GmbH（NXAI公司） ； ELLIS Unit, LIT AI Lab, Institute for Machine Learning, JKU Linz（ELLIS单元、LIT AI实验室、机器学习研究所） ； Interdisciplinary Transformation University Austria（跨学科转型大学奥地利） 纠错

Comments presented at NeurIPS 2025

Diffusion Classifiers Understand Compositionality, but Conditions Apply

Yujin Jeong, Arnas Uselis, Seong Joon Oh, Anna Rohrbach

机构 * TU Darmstadt（图恩大学） ； University of Tübingen（图宾根大学） 纠错

Comments NeurIPS 2025 Datasets and Benchmarks

Words That Unite The World: A Unified Framework for Deciphering Central Bank Communications Globally

Agam Shah, Siddhant Sukhani, Huzaifa Pardawala, Saketh Budideti, Riya Bhadani, Rudra Gopal, Siddhartha Somani, Rutwik Routu, Michael Galarnyk, Soungmin Lee, Arnav Hiray, Akshar Ravichandran, Eric Kim, Pranav Aluru, Joshua Zhang, Sebastian Jaskowski, Veer Guda, Meghaj Tarte, Liqin Ye, Spencer Gosden, Rachel Yuh, Sloka Chava, Sahasra Chava, Dylan Patrick Kelly, Aiden Chiang, Harsit Mittal, Sudheer Chava

Comments Accepted at NeurIPS 2025 (main conference)

Multi-head Temporal Latent Attention

Keqi Deng, Philip C. Woodland

机构 * Department of Engineering, University of Cambridge（工程系，剑桥大学） 纠错

Comments Accepted by NeurIPS 2025

Reasoning by Superposition: A Theoretical Perspective on Chain of Continuous Thought

Hanlin Zhu, Shibo Hao, Zhiting Hu, Jiantao Jiao, Stuart Russell, Yuandong Tian

机构 * UC Berkeley（伯克利大学） ； UCSD（加州大学圣达菲分校） ； Meta AI 纠错

Comments 26 pages, 7 figures, NeurIPS 2025

Characterization and Learning of Causal Graphs from Hard Interventions

Zihan Zhou, Muhammad Qasim Elahi, Murat Kocaoglu

机构 * Johns Hopkins University（约翰霍普金斯大学） ； Purdue University（普渡大学） 纠错

Comments Accepted at NeurIPS 2025

PolyMath: Evaluating Mathematical Reasoning in Multilingual Contexts

Yiming Wang, Pei Zhang, Jialong Tang, Haoran Wei, Baosong Yang, Rui Wang, Chenshu Sun, Feitong Sun, Jiran Zhang, Junxuan Wu, Qiqian Cang, Yichang Zhang, Fei Huang, Junyang Lin, Fei Huang, Jingren Zhou

机构 * Qwen Team, Alibaba Group（阿里集团Qwen团队） ； Shanghai Jiao Tong University（上海交通大学） 纠错

Comments Accepted by NeurIPS 2025

Hardware-aligned Hierarchical Sparse Attention for Efficient Long-term Memory Access

Xiang Hu, Jiaqi Leng, Jun Zhao, Kewei Tu, Wei Wu

机构 * Ant Group（蚂蚁集团） ； Fudan University（复旦大学） ； ShanghaiTech University（上海科技大学） 纠错

Comments Accepted to NeurIPS 2025

Sampling-Efficient Test-Time Scaling: Self-Estimating the Best-of-N Sampling in Early Decoding

Yiming Wang, Pei Zhang, Siyuan Huang, Baosong Yang, Zhuosheng Zhang, Fei Huang, Rui Wang

机构 * School of Computer Science, Shanghai Jiao Tong University（上海交通大学计算机科学学院） ； Tongyi Lab, Alibaba Group（阿里集团通义实验室） 纠错

Comments Accepted by NeurIPS 2025 (Spotlight)

Split Gibbs Discrete Diffusion Posterior Sampling

Wenda Chu, Zihui Wu, Yifan Chen, Yang Song, Yisong Yue

机构 * California Institute of Technology（加州理工学院） ； New York University（纽约大学） ； OpenAI 纠错

Comments Accepted to NeurIPS 2025

LLM Strategic Reasoning: Agentic Study through Behavioral Game Theory

Jingru Jia, Zehua Yuan, Junhao Pan, Paul E. McNamara, Deming Chen

机构 * University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） 纠错

Comments Accepted by NeurIPS 2025

Auto-Search and Refinement: An Automated Framework for Gender Bias Mitigation in Large Language Models

Yue Xu, Chengyan Fu, Li Xiong, Sibei Yang, Wenjie Wang

机构 * School of Information Science and Technology, ShanghaiTech University（信息科学与技术学院，上海科技大学） ； Emory University（埃默里大学） ； Sun Yat-sen University（中山大学） 纠错

Comments Accepted to NeurIPS 2025

FIPER: Factorized Features for Robust Image Super-Resolution and Compression

Yang-Che Sun, Cheng Yu Yeo, Ernie Chu, Jun-Cheng Chen, Yu-Lun Liu

机构 * National Yang Ming Chiao Tung University（国家阳明交通大学） ； Johns Hopkins University（约翰霍普金斯大学） ； Academia Sinica（中国科学院） 纠错

Comments NeurIPS 2025. Project page: https://jayisaking.github.io/FIPER/

Gymnasium: A Standard Interface for Reinforcement Learning Environments

Mark Towers, Ariel Kwiatkowski, Jordan Terry, John U. Balis, Gianluca De Cola, Tristan Deleu, Manuel Goulão, Andreas Kallinteris, Markus Krimmel, Arjun KG, Rodrigo Perez-Vicente, Andrea Pierré, Sander Schulhoff, Jun Jet Tai, Hannah Tan, Omar G. Younis

机构 * University of Southampton（南安普顿大学） ； Farama Foundation（Farama基金会） ； Meta AI, FAIR（Meta AI 和 FAIR） ； MILA, Université de Montréal（蒙特利尔大学 MILA 实验室） ； NeuralShift ； Technical University of Crete（希腊技术大学） ； EarthBrain ； UMass Lowell（马萨诸塞大学洛厄尔分校） ； University of Maryland（马里兰大学） ； Coventry University（科文特花园大学） 纠错

Comments Accepted at NeurIPS Datasets and Benchmarks 2025

Knolling Bot: Teaching Robots the Human Notion of Tidiness

Yuhang Hu, Judah Goldfeder, Zhizhuo Zhang, Xinyue Zhu, Ruibo Liu, Philippe Wyder, Jiong Lin, Hod Lipson

机构 * Columbia University（哥伦比亚大学） 纠错

Comments Accepted at the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Creative AI Track

On the Variance, Admissibility, and Stability of Empirical Risk Minimization

Gil Kur, Eli Putterman, Alexander Rakhlin

Comments Appeared at Neurips 2023 (spotlight)

Probing Knowledge Holes in Unlearned LLMs

Myeongseob Ko, Hoang Anh Just, Charles Fleming, Ming Jin, Ruoxi Jia

机构 * Virginia Tech（弗吉尼亚理工大学） ； Cisco Research（思科研究） 纠错

Comments The Thirty-ninth Annual Conference on Neural Information Processing Systems

SpecAttn: Speculating Sparse Attention

Harsh Shah

机构 * Machine Learning Department（机器学习系） ； Carnegie Mellon University（卡内基梅隆大学） 纠错

Comments Accepted to NeurIPS 2025 Workshop on Structured Probabilistic Inference & Generative Modeling

Leveraging Generic Time Series Foundation Models for EEG Classification

Théo Gnassounou, Yessin Moakher, Shifeng Xie, Vasilii Feofanov, Ievgen Redko

机构 * Université Paris-Saclay, Inria, CEA（巴黎萨克雷大学、法国国家信息与自动化研究所、法国原子能委员会） ； Huawei Noah’s Ark Lab（华为诺亚实验室） 纠错

Journal ref NeurIPS 2025 Workshop "Recent Advances in Time Series Foundation Models Have We Reached the 'BERT Moment'?"

Learning Sparse Approximate Inverse Preconditioners for Conjugate Gradient Solvers on GPUs

Zherui Yang, Zhehao Li, Kangbo Lyu, Yixuan Li, Tao Du, Ligang Liu

机构 * University of Science and Technology of China（中国科学技术大学） ； Stanford University（斯坦福大学） ； Tsinghua University（清华大学） ； Shanghai Qi Zhi Institute（上海启智研究院） 纠错

Comments NeurIPS 2025, poster

NAUTILUS: A Large Multimodal Model for Underwater Scene Understanding

Wei Xu, Cheng Wang, Dingkang Liang, Zongchuang Zhao, Xingyu Jiang, Peng Zhang, Xiang Bai

机构 * National University of Defense Technology（国防科技大学） 纠错

Comments Accepted to NeurIPS 2025. Data and models are available at https://github.com/H-EmbodVis/NAUTILUS