NeurIPS

Less is More: Local Intrinsic Dimensions of Contextual Language Models

Benjamin Matthias Ruppik, Julius von Rohrscheidt, Carel van Niekerk, Michael Heck, Renato Vukovic, Shutong Feng, Hsien-chin Lin, Nurul Lubis, Bastian Rieck, Marcus Zibrowius, Milica Gašić

机构 * Faculty of Mathematics and Natural Sciences, Heinrich Heine University Düsseldorf（明斯特大学数学与自然科学学院） ； Institute of AI for Health, Helmholtz Munich（健康人工智能研究所） ； Technical University of Munich（慕尼黑技术大学） ； University of Fribourg（弗里堡大学） 纠错

Comments Accepted at the 39th Conference on Neural Information Processing Systems (NeurIPS 2025; in press). 10 pages, with an additional 17 pages in the appendix. Our code is available at https://github.com/aidos-lab/Topo_LLM_public and https://github.com/aidos-lab/grokking-via-lid

Journal ref Advances in Neural Information Processing Systems, Volume 38 (NeurIPS 2025)

ProSpero: Active Learning for Robust Protein Design Beyond Wild-Type Neighborhoods

Michal Kmicikiewicz, Vincent Fortuin, Ewa Szczurek

机构 * Institute of AI for Health, Helmholtz Munich（亥姆霍兹慕尼黑人工智能健康研究所） ； School of Computation, Information and Technology, Technical University of Munich（慕尼黑技术大学计算、信息与技术学院） ； Helmholtz AI（亥姆霍兹人工智能研究所） ； Munich Center for Machine Learning（慕尼黑机器学习中心） ； Faculty of Mathematics, Informatics and Mechanics, University of Warsaw（华沙大学数学、信息学与力学学院） 纠错

Comments NeurIPS 2025

MOOSE-Chem2: Exploring LLM Limits in Fine-Grained Scientific Hypothesis Discovery via Hierarchical Search

Zonglin Yang, Wanhao Liu, Ben Gao, Yujie Liu, Wei Li, Tong Xie, Lidong Bing, Wanli Ouyang, Erik Cambria, Dongzhan Zhou

机构 * Nanyang Technological University（南洋理工大学） ； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室） ； University of Science and Technology of China（中国科学技术大学） ； Wuhan University（武汉大学） ； National University of Singapore（新加坡国立大学） ； University of New South Wales（新南威尔士大学） ； MiroMind ； The Chinese University of Hong Kong（香港中文大学） 纠错

Comments Accepted by NeurIPS 2025

DataRater: Meta-Learned Dataset Curation

Dan A. Calian, Gregory Farquhar, Iurii Kemaev, Luisa M. Zintgraf, Matteo Hessel, Jeremy Shar, Junhyuk Oh, András György, Tom Schaul, Jeffrey Dean, Hado van Hasselt, David Silver

机构 * Google（谷歌） ； DeepMind 纠错

Comments NeurIPS 2025

The Atlas of In-Context Learning: How Attention Heads Shape In-Context Retrieval Augmentation

Patrick Kahardipraja, Reduan Achtibat, Thomas Wiegand, Wojciech Samek, Sebastian Lapuschkin

机构 * Department of Artificial Intelligence, Fraunhofer Heinrich Hertz Institute（人工智能系，弗劳恩霍夫海因里希·赫兹研究所） ； Department of Electrical Engineering and Computer Science, Technische Universität Berlin（电气工程与计算机科学系，柏林技术大学） ； BIFOLD - Berlin Institute for the Foundations of Learning and Data（柏林学习与数据基础研究所） ； Centre of eXplainable Artificial Intelligence, Technological University Dublin（可解释人工智能中心，都柏林技术大学） 纠错

Comments Accepted at NeurIPS 2025

Fast Rate Bounds for Multi-Task and Meta-Learning with Different Sample Sizes

Hossein Zakerinia, Christoph H. Lampert

机构 * Institute of Science and Technology Austria (ISTA)（奥地利科学与技术研究院） 纠错

Comments Conference on Neural Information Processing Systems (NeurIPS), 2025

Validating LLM-as-a-Judge Systems under Rating Indeterminacy

Luke Guerdan, Solon Barocas, Kenneth Holstein, Hanna Wallach, Zhiwei Steven Wu, Alexandra Chouldechova

Comments NeurIPS 2025

Convolutional Hierarchical Deep Learning Neural Networks-Tensor Decomposition (C-HiDeNN-TD): a scalable surrogate modeling approach for large-scale physical systems

Jiachen Guo, Chanwook Park, Xiaoyu Xie, Zhongsheng Sang, Gregory J. Wagner, Wing Kam Liu

Journal ref NeurIPS 2024 workshop on Data-driven and Differentiable Simulations, Surrogates, and Solvers

Increasing LLM Coding Capabilities through Diverse Synthetic Coding Tasks

Amal Abed, Ivan Lukic, Jörg K. H. Franke, Frank Hutter

机构 * University of Freiburg（弗赖堡大学） ； ELLIS Institute Tübingen（图宾根ELLIS研究所） ； Open-Sci Collective（开放科学集体） ； LAION ； Prior Labs（Prior实验室） 纠错

Comments Presented at the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: The 4th Deep Learning for Code Workshop (DL4C)

Evaluation of Vision-LLMs in Surveillance Video

Pascal Benschop, Cristian Meo, Justin Dauwels, Jelte P. Mense

机构 * Department of Computer Science（计算机科学系） ； Delft University of Technology（代尔夫特理工大学） ； LatentWorlds AI ； National Policelab AI & Model-Driven Decisions Lab（国家警务实验室AI与模型驱动决策实验室） 纠错

Comments Accepted as poster in the NeurIPS 2025 Workshop on Space in Vision, Language, and Embodied AI

Implicit Modeling for Transferability Estimation of Vision Foundation Models

Yaoyan Zheng, Huiqun Wang, Nan Zhou, Di Huang

机构 * State Key Laboratory of Complex and Critical Software Environment, Beihang University, Beijing, China（复杂与关键软件环境国家重点实验室，北京航空航天大学） ； School of Computer Science and Engineering, Beihang University, Beijing, China（计算机科学与工程学院，北京航空航天大学） 纠错

Comments Accepted by NeurIPS 2025

Beyond Higher Rank: Token-wise Input-Output Projections for Efficient Low-Rank Adaptation

Shiwei Li, Xiandi Luo, Haozhao Wang, Xing Tang, Ziqiang Cui, Dugang Liu, Yuhua Li, Xiuqiang He, Ruixuan Li

机构 * Huazhong University of Science and Technology（华中科技大学） ； Shenzhen Technology University（深圳技术大学） ； City University of Hong Kong（香港城市大学） ； Shenzhen University（深圳大学） 纠错

Comments Accepted by NeurIPS 2025

Think before Recommendation: Autonomous Reasoning-enhanced Recommender

Xiaoyu Kong, Junguang Jiang, Bin Liu, Ziru Xu, Han Zhu, Jian Xu, Bo Zheng, Jiancan Wu, Xiang Wang

机构 * Taobao & Tmall Group of Alibaba, China（阿里巴巴集团淘宝与天猫事业部） ； National University of Singapore（新加坡国立大学） ； Institute of Dataspace, Hefei Comprehensive National Science Center（合肥综合国家科学中心数据空间研究院） ； Shanghai Key Laboratory of Data Science（上海数据科学重点实验室） 纠错

Comments NeurIPS 2025 poster

LLM Meets Diffusion: A Hybrid Framework for Crystal Material Generation

Subhojyoti Khastagir, Kishalay Das, Pawan Goyal, Seung-Cheol Lee, Satadeep Bhattacharjee, Niloy Ganguly

机构 * Indian Institute of Technology, Kharagpur, India（印度理工学院，克哈格普尔分校） ； Indo Korea Science and Technology Center, Bangalore, India（印韩科学技术中心，班加罗尔） 纠错

Comments NeurIPS 2025

LoMix: Learnable Weighted Multi-Scale Logits Mixing for Medical Image Segmentation

Md Mostafijur Rahman, Radu Marculescu

机构 * Department of Electrical and Computer Engineering（电气与计算机工程系） ； The University of Texas at Austin（德克萨斯大学奥斯汀分校） 纠错

Comments 25 pages, 13 figures, NeurIPS 2025 accepted paper

SceneDecorator: Towards Scene-Oriented Story Generation with Scene Planning and Scene Consistency

Quanjian Song, Donghao Zhou, Jingyu Lin, Fei Shen, Jiaze Wang, Xiaowei Hu, Cunjian Chen, Pheng-Ann Heng

机构 * Monash University（莫纳什大学） ； The Chinese University of Hong Kong（香港中文大学） ； National University of Singapore（新加坡国立大学） ； South China University of Technology（华南理工大学） 纠错

Comments Accepted by NeurIPS 2025; Project Page: https://lulupig12138.github.io/SceneDecorator

Exploring Semantic-constrained Adversarial Example with Instruction Uncertainty Reduction

Jin Hu, Jiakai Wang, Linna Jing, Haolin Li, Haodong Liu, Haotong Qin, Aishan Liu, Ke Xu, Xianglong Liu

机构 * State Key Laboratory of Complex & Critical Software Environment (CCSE), Beihang University（复杂与关键软件环境国家重点实验室，北京航空航天大学） ； Zhongguancun Laboratory（中关村实验室） ； School of Computer Science and Engineering, Beihang University（北京航空航天大学计算机科学与工程学院） ； Department of Information Technology and Electrical Engineering, ETH Zurich（苏黎世联邦理工学院信息科技与电气工程系） 纠错

Comments NeurIPS 2025

Artificial Hivemind: The Open-Ended Homogeneity of Language Models (and Beyond)

Liwei Jiang, Yuanjun Chai, Margaret Li, Mickel Liu, Raymond Fok, Nouha Dziri, Yulia Tsvetkov, Maarten Sap, Alon Albalak, Yejin Choi

Comments NeurIPS 2025 D&B Paper (Oral); Camera-Ready Version

Learning Reconfigurable Representations for Multimodal Federated Learning with Missing Data

Duong M. Nguyen, Trong Nghia Hoang, Thanh Trung Huynh, Quoc Viet Hung Nguyen, Phi Le Nguyen

机构 * University of Illinois Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； Washington State University（华盛顿州立大学） ； VinUniversity（文大学） ； Griffin University（格里芬大学） ； Hanoi University of Science and Technology（河内科学技术大学） 纠错

Comments Accepted at NeurIPS 2025

Encoder-Decoder Diffusion Language Models for Efficient Training and Inference

Marianne Arriola, Yair Schiff, Hao Phung, Aaron Gokaslan, Volodymyr Kuleshov

机构 * Department of Computer Science, Cornell University（计算机科学系，康奈尔大学） 纠错

Comments NeurIPS 2025. We provide the code at https://github.com/kuleshov-group/e2d2

Semantic Surgery: Zero-Shot Concept Erasure in Diffusion Models

Lexiang Xiong, Chengyu Liu, Jingwen Ye, Yan Liu, Yuecong Xu

机构 * National University of Singapore（国立新加坡大学） ； Sichuan University（四川大学） 纠错

Comments Accepted to the 39th Conference on Neural Information Processing Systems (NeurIPS 2025). Code is available at https://github.com/Lexiang-Xiong/Semantic-Surgery

Once Upon an Input: Reasoning via Per-Instance Program Synthesis

Adam Stein, Neelay Velingker, Mayur Naik, Eric Wong

机构 * University of Pennsylvania（宾夕法尼亚大学） 纠错

Comments Accepted at NeurIPS 2025. 34 pages, 7 figures

FOXES: A Framework For Operational X-ray Emission Synthesis

Griffin T. Goodwin, Jayant Biradar, Alison J. March, Christoph Schirninger, Robert Jarolim, Angelos Vourlidas, Lorien Pratt

Comments Accepted to Machine Learning and the Physical Sciences Workshop, NeurIPS 2025

Inductive Transfer Learning for Graph-Based Recommenders

Florian Grötschla, Elia Trachsel, Luca A. Lanzendörfer, Roger Wattenhofer

机构 * ETH Zurich（苏黎世联邦理工学院） 纠错

Comments Accepted at the New Perspectives in Graph Machine Learning Workshop at NeurIPS 2025

SAO-Instruct: Free-form Audio Editing using Natural Language Instructions

Michael Ungersböck, Florian Grötschla, Luca A. Lanzendörfer, June Young Yi, Changho Choi, Roger Wattenhofer

机构 * ETH Zurich（苏黎世联邦理工学院） 纠错

Comments Accepted at NeurIPS 2025

Windsock is Dancing: Adaptive Multimodal Retrieval-Augmented Generation

Shu Zhao, Tianyi Shen, Nilesh Ahuja, Omesh Tickoo, Vijaykrishnan Narayanan

机构 * The Pennsylvania State University（宾夕法尼亚州立大学） ； Intel（英特尔） 纠错

Comments Accepted at NeurIPS 2025 UniReps Workshop

Alias-Free ViT: Fractional Shift Invariance via Linear Attention

Hagay Michaeli, Daniel Soudry

机构 * Technion（技术离子） 纠错

Comments Accepted at NeurIPS 2025. Code is available at https://github.com/hmichaeli/alias_free_vit

If You Want to Be Robust, Be Wary of Initialization

Sofiane Ennadir, Johannes F. Lutzeyer, Michalis Vazirgiannis, El Houcine Bergou

机构 * KTH Stockholm, Sweden（瑞典斯德哥尔摩皇家理工学院） ； LIX, Ecole Polytechnique IP Paris, France（法国巴黎高等理工学院LIX研究所） ； KTH & Ecole Polytechnique Stockholm, Sweden（瑞典斯德哥尔摩皇家理工学院及巴黎高等理工学院） ； UM6P Benguerir, Morocco（摩洛哥本格里大学） 纠错

Comments Accepted at NeurIPS 2024

Enhancing Graph Classification Robustness with Singular Pooling

Sofiane Ennadir, Oleg Smirnov, Yassine Abbahaddou, Lele Cao, Johannes F. Lutzeyer

机构 * King AI Labs, Microsoft Gaming（King AI实验室，微软游戏） ； LIX, École Polytechnique, IP Paris（LIX，巴黎高等理工学院，IP巴黎） 纠错

Comments Accepted at Neurips 2025

Prediction-Powered Semi-Supervised Learning with Online Power Tuning

Noa Shoham, Ron Dorfman, Shalev Shaer, Kfir Y. Levy, Yaniv Romano

机构 * Department of Electrical and Computer Engineering, Technion IIT（电气与计算机工程系，技术离子研究所） ； Department of Computer Science, Technion IIT（计算机科学系，技术离子研究所） 纠错

Comments NeurIPS 2025