NeurIPS

TangledFeatures: Robust Feature Selection in Highly Correlated Spaces

Allen Daniel Sunny

机构 * University of Maryland, College Park（马里兰大学学院公园分校） 纠错

Comments Accepted for poster presentation at the Machine Learning for Structural Biology (MLSB) Workshop @ NeurIPS 2025, co-located with NeurIPS 2025 (San Diego, USA). Non-archival

Exploring and Leveraging Class Vectors for Classifier Editing

Jaeik Kim, Jaeyoung Do

机构 * AIDAS Laboratory（AIDAS实验室） ； IPAI ； ECE（电子工程系） ； Seoul National University（首尔国立大学） 纠错

Comments Accepted in NeurIPS 2025

Normal-Abnormal Guided Generalist Anomaly Detection

Yuexin Wang, Xiaolei Wang, Yizheng Gong, Jimin Xiao

机构 * Xi’an Jiaotong-Liverpool University（西安交通大学利物浦大学） ； University of Liverpool（利物浦大学） 纠错

Comments Accepted by NeurIPS 2025

Ascent Fails to Forget

Ioannis Mavrothalassitis, Pol Puigdemont, Noam Itzhak Levi, Volkan Cevher

机构 * LIONS, École Polytechnique Fédérale de Lausanne (EPFL), Lausanne, Switzerland（洛桑联邦理工学院LIONS研究所） 纠错

Comments NeurIPS 2025

Towards Reliable and Holistic Visual In-Context Learning Prompt Selection

Wenxiao Wu, Jing-Hao Xue, Chengming Xu, Chen Liu, Xinwei Sun, Changxin Gao, Nong Sang, Yanwei Fu

机构 * Huazhong University of Science and Technology（华中科技大学） ； Shanghai Innovation Institute（上海创新研究院） ； University College London（伦敦大学学院） ； Tencent Youtu Lab（腾讯优图实验室） ； The Hong Kong University of Science and Technology（香港理工大学） ； Fudan University（复旦大学） 纠错

Comments Accepted by NeurIPS 2025

F-Adapter: Frequency-Adaptive Parameter-Efficient Fine-Tuning in Scientific Machine Learning

Hangwei Zhang, Chun Kang, Yan Wang, Difan Zou

机构 * School of Computing and Data Science, The University of Hong Kong（计算与数据科学学院，香港大学） ； Institute for AI Industry Research, Tsinghua University（人工智能产业研究院，清华大学） ； Beihang University（北航大学） 纠错

Comments NeurIPS 2025 Main Track

Graph-based Neural Space Weather Forecasting

Daniel Holmberg, Ivan Zaitsev, Markku Alho, Ioanna Bouri, Fanni Franssila, Haewon Jeong, Minna Palmroth, Teemu Roos

机构 * University of Helsinki（赫尔辛基大学） ； UC Santa Barbara（圣巴巴拉大学） ； FMI Space and Earth Observation Centre（FMI空间与地球观测中心） 纠错

Comments 20 pages, 18 figures. Accepted to the NeurIPS 2025 Workshop on Machine Learning and the Physical Sciences

Migration as a Probe: A Generalizable Benchmark Framework for Specialist vs. Generalist Machine-Learned Force Fields

Yi Cao, Paulette Clancy

机构 * Johns Hopkins University（约翰霍普金斯大学） 纠错

Comments 10 pages (+Appendix 22 pages), 8 figures. To appear in 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: AI for Accelerated Materials Discovery (AI4Mat) Workshop. Code and datasets available at https://github.com/yicao-elina/MigrationBench.git

Hydra: A Modular Architecture for Efficient Long-Context Reasoning

Siddharth Chaudhary, Dev Patel, Maheep Chaudhary, Bennett Browning

机构 * St Paul’s School, London（伦敦圣保罗学校） ； University of California, Berkeley（加州大学伯克利分校） 纠错

Comments Updated with the new paper accepted to NeurIPS workshop

Scaling Physical Reasoning with the PHYSICS Dataset

Shenghe Zheng, Qianjia Cheng, Junchi Yao, Mengsong Wu, Haonan He, Ning Ding, Yu Cheng, Shuyue Hu, Lei Bai, Dongzhan Zhou, Ganqu Cui, Peng Ye

机构 * Shanghai AI Laboratory（上海人工智能实验室） ； Harbin Institute of Technology（哈尔滨工业大学） ； Tsinghua University（清华大学） ； CUHK（香港中文大学） ； BUAA（北京航空航天大学） ； UESTC（电子科技大学） ； Soochow University（苏州大学） ； USTC（中国科学技术大学） 纠错

Comments Accepted to the NeurIPS Datasets and Benchmarks Track

Hyperbolic Dataset Distillation

Wenyuan Li, Guang Li, Keisuke Maeda, Takahiro Ogawa, Miki Haseyama

机构 * Hokkaido University（北海道大学） 纠错

Comments Accepted to NeurIPS 2025

msf-CNN: Patch-based Multi-Stage Fusion with Convolutional Neural Networks for TinyML

Zhaolan Huang, Emmanuel Baccelli

机构 * Freie Universität Berlin（柏林自由大学） ； Einstein Center Digital Future Inria, France（数字未来埃因斯坦中心，Inria，法国） 纠错

Comments NeurIPS 2025 (poster)

SPICE: A Synergistic, Precise, Iterative, and Customizable Image Editing Workflow

Kenan Tang, Yanhong Li, Yao Qin

机构 * University of California, Santa Barbara（加州大学圣芭芭拉分校） 纠错

Comments The paper has been accepted to NeurIPS Creative AI Track 2025. Figure 4(c) has been accepted to CVPR AI Art Gallery 2025

LinEAS: End-to-end Learning of Activation Steering with a Distributional Loss

Pau Rodriguez, Michal Klein, Eleonora Gualdoni, Valentino Maiorca, Arno Blaas, Luca Zappella, Marco Cuturi, Xavier Suau

机构 * Apple（苹果公司） ； Sapienza（拉维尼恩大学） 纠错

Comments NeurIPS 2025

PrefixKV: Adaptive Prefix KV Cache is What Vision Instruction-Following Models Need for Efficient Generation

Ao Wang, Hui Chen, Jiaxin Li, Jianchao Tan, Kefeng Zhang, Xunliang Cai, Zijia Lin, Jungong Han, Guiguang Ding

机构 * School of Software, Tsinghua University（清华大学软件学院） ； BNRist, Tsinghua University（清华大学BNRist） ； Meituan Inc.（美团公司） ； Department of Automation, Tsinghua University（清华大学自动化系） 纠错

Comments NeurIPS 2025 Camera-ready Version

RDD: Retrieval-Based Demonstration Decomposer for Planner Alignment in Long-Horizon Tasks

Mingxuan Yan, Yuping Wang, Zechun Liu, Jiachen Li

机构 * University of California, Riverside（加州大学河滨分校） ； University of Michigan（密歇根大学） ； Meta AI 纠错

Comments 39th Conference on Neural Information Processing Systems (NeurIPS 2025); Project Website: rdd-neurips.github.io

The ML.ENERGY Benchmark: Toward Automated Inference Energy Measurement and Optimization

Jae-Won Chung, Jeff J. Ma, Ruofan Wu, Jiachen Liu, Oh Jun Kweon, Yuxuan Xia, Zhiyu Wu, Mosharaf Chowdhury

机构 * University of Michigan（密歇根大学） 纠错

Comments NeurIPS D&B 2025 (Spotlight) | Benchmark: https://github.com/ml-energy/benchmark | Leaderboard: https://ml.energy/leaderboard

Strategyproof Reinforcement Learning from Human Feedback

Thomas Kleine Buening, Jiarui Gan, Debmalya Mandal, Marta Kwiatkowska

机构 * ETH Zurich（苏黎世联邦理工学院） ； ETH AI Center（ETH人工智能中心） ； Department of Computer Science University of Oxford（牛津大学计算机科学系） ； Department of Computer Science University of Warwick（Warwick大学计算机科学系） 纠错

Comments To appear at NeurIPS 2025

LinPrim: Linear Primitives for Differentiable Volumetric Rendering

Nicolas von Lützow, Matthias Nießner

机构 * Technical University of Munich（慕尼黑技术大学） 纠错

Comments Project page: https://nicolasvonluetzow.github.io/LinPrim - Project video: https://youtu.be/NRRlmFZj5KQ - Accepted at NeurIPS 2025

Agentic NL2SQL to Reduce Computational Costs

Dominik Jehle, Lennart Purucker, Frank Hutter

机构 * University of Freiburg（弗赖堡大学） ； Prior Labs（Prior公司） ； ELLIS Institute（ELLIS研究所） ； Tübingen University of Freiburg（弗赖堡大学图宾根分校） 纠错

Comments Accepted at the NeurIPS 2025 Workshop on Efficient Reasoning. 10 pages, 11 figures

Galaxy Morphology Classification with Counterfactual Explanation

Zhuo Cao, Lena Krieger, Hanno Scharr, Ira Assent

机构 * IAS-8, Forschungszentrum Jülich, Germany（贾斯特中心Jülich德国第8研究所） ； Aarhus University, Denmark（奥胡斯大学丹麦） 纠错

Comments Accepted to the Machine Learning and the Physical Sciences Workshop at NeurIPS 2024 (non-archival)

Redundancy-Aware Test-Time Graph Out-of-Distribution Detection

Yue Hou, He Zhu, Ruomei Liu, Yingke Su, Junran Wu, Ke Xu

机构 * State Key Laboratory of Complex & Critical Software Environment（复杂与关键软件环境国家重点实验室） ； Shen Yuan Honors College（申元荣誉学院） 纠错

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

Eyes Wide Open: Ego Proactive Video-LLM for Streaming Video

Yulin Zhang, Cheng Shi, Yang Wang, Sibei Yang

机构 * ShanghaiTech University（上海科技大学） ； School of Computer Science and Engineering, Sun Yat-sen University（中山大学计算机科学与工程学院） ； School of Computing and Data Science, The University of Hong Kong（香港大学计算科学与数据科学学院） 纠错

Comments Accepted at NeurIPS 2025 (preview; camera-ready in preparation)

Deep Compositional Phase Diffusion for Long Motion Sequence Generation

Ho Yin Au, Jie Chen, Junkun Jiang, Jingyu Xiang

机构 * Hong Kong Baptist University（香港 Baptist 大学） 纠错

Comments Accepted by NeurIPS 2025 (Oral)

Stop-RAG: Value-Based Retrieval Control for Iterative RAG

Jaewan Park, Solbee Cho, Jay-Yoon Lee

机构 * Seoul National University（首尔国立大学） 纠错

Comments NeurIPS 2025 MTI-LLM Workshop

Hierarchical Simulation-Based Inference of Supernova Power Sources and their Physical Properties

Edgar P. Vidal, Alexander T. Gagliano, Carolina Cuesta-Lazaro

Comments 9 pages, 4 figures, Accepted at the 2025 Machine Learning and the Physical Sciences (ML4PS) workshop at NeurIPS

Vgent: Graph-based Retrieval-Reasoning-Augmented Generation For Long Video Understanding

Xiaoqian Shen, Wenxuan Zhang, Jun Chen, Mohamed Elhoseiny

机构 * King Abdullah University of Science and Technology（卡布斯大学） 纠错

Comments NeurIPS 2025 (Spotlight). Webpage at https://xiaoqian-shen.github.io/Vgent

Robust or Suggestible? Exploring Non-Clinical Induction in LLM Drug-Safety Decisions

Siying Liu, Shisheng Zhang, Indu Bala

Comments Preprint of a paper accepted as a poster at the NeurIPS 2025 Workshop on Generative AI for Health (GenAI4Health). The final camera-ready workshop version may differ. Licensed under CC BY 4.0

Weight Weaving: Parameter Pooling for Data-Free Model Merging

Levy Chaves, Eduardo Valle, Sandra Avila

机构 * Recod.ai Lab.（Recod.ai 实验室） ； Instituto de Computação Universidade Estadual de Campinas (UNICAMP)（计算研究所，坎皮纳斯州立大学） 纠错

Comments 17 pages, 3 figures. Accepted at the 3rd UniReps Workshop @ NeurIPS 2025

Self-Training with Dynamic Weighting for Robust Gradual Domain Adaptation

Zixi Wang, Yushe Cao, Yubo Huang, Jinzhu Wei, Jingzehua Xu, Shuai Zhang, Xin Lai

机构 * University of Electronic Science and Technology of China（电子科技大学） ； Tsinghua University（清华大学） ； Zhenguan AI Lab ； Southwest Jiaotong University（西南交通大学） ； Shanghai University（上海大学） ； New Jersey Institute of Technology（新泽西理工学院） 纠错

Comments It had formerly appeared as arXiv:2501.19159v2 in error. Accepted by NIPS 25