NeurIPS

The Last Vote: A Multi-Stakeholder Framework for Language Model Governance

Subramanyam Sahoo, Aditi Chhawacharia

Comments This paper has been accepted to the NeurIPS 2025 Workshop on Algorithmic Collective Action (ACA@NeurIPS 2025). The submission is 26 pages including the appendix and includes the NeurIPS checklist. A big thanks to Avijit Ghosh

Natural gradient descent for improving variational inference based classification of radio galaxies

Devina Mohan, Anna M. M. Scaife

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

TokenSqueeze: Performance-Preserving Compression for Reasoning LLMs

Yuxiang Zhang, Zhengxu Yu, Weihang Pan, Zhongming Jin, Qiang Fu, Deng Cai, Binbin Lin, Jieping Ye

机构 * State Key Lab of CAD&CG, Zhejiang University（浙江大学CAD&CG国家重点实验室） ； School of Software Technology, Zhejiang University（浙江大学软件学院） ； Alibaba Cloud（阿里云） ； Zhiyuan Research Institute（智源研究院） 纠错

Comments Accepted to NeurIPS 2025

Toward Explainable Offline RL: Analyzing Representations in Intrinsically Motivated Decision Transformers

Leonardo Guiducci, Antonio Rizzo, Giovanna Maria Dimitri

机构 * DISPOC University of Siena（DISPOC 硅恩纳大学） ； University of Siena（硅恩纳大学） ； University of Milan（米兰大学） 纠错

Comments Accepted for poster presentation at the NeurIPS 2025 workshop "CogInterp: Interpreting Cognition in Deep Learning Models", San Diego, CA, USA

Compress, Gather, and Recompute: REFORMing Long-Context Processing in Transformers

Woomin Song, Sai Muralidhar Jayanthi, Srikanth Ronanki, Kanthashree Mysore Sathyendra, Jinwoo Shin, Aram Galstyan, Shubham Katiyar, Sravan Babu Bodapati

机构 * KAIST（韩国科学技术院） ； Amazon AGI（亚马逊人工智能实验室） 纠错

Comments NeurIPS 2025

ZPressor: Bottleneck-Aware Compression for Scalable Feed-Forward 3DGS

Weijie Wang, Donny Y. Chen, Zeyu Zhang, Duochao Shi, Akide Liu, Bohan Zhuang

机构 * ZIP Lab, Zhejiang University（浙江大学ZIP实验室） ； Monash University（墨尔本大学） 纠错

Comments NeurIPS 2025, Project Page: https://lhmd.top/zpressor, Code: https://github.com/ziplab/ZPressor

Meta-Learning an In-Context Transformer Model of Human Higher Visual Cortex

Muquan Yu, Mu Nan, Hossein Adeli, Jacob S. Prince, John A. Pyles, Leila Wehbe, Margaret M. Henderson, Michael J. Tarr, Andrew F. Luo

Comments Accepted to NeurIPS 2025. Website: https://github.com/leomqyu/BraInCoRL

Hogwild! Inference: Parallel LLM Generation via Concurrent Attention

Gleb Rodionov, Roman Garipov, Alina Shutova, George Yakushev, Erik Schultheis, Vage Egiazarian, Anton Sinitsin, Denis Kuznedelev, Dan Alistarh

机构 * Yandex ； HSE University（俄罗斯高等经济大学） ； IST Austria（国际研究所） 纠错

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

A is for Absorption: Studying Feature Splitting and Absorption in Sparse Autoencoders

David Chanin, James Wilken-Smith, Tomáš Dulka, Hardik Bhatnagar, Satvik Golechha, Joseph Bloom

机构 * LASR Labs（LASR实验室） ； University College London（伦敦大学学院） ； Tübingen AI Center, University of Tübingen（图宾根大学人工智能中心） ； MATS ； Decode Research（Decode研究） 纠错

Comments Accepted at NeurIPS 2025 (Oral)

Learning Time-Scale Invariant Population-Level Neural Representations

Eshani Patel, Yisong Yue, Geeling Chau

机构 * Computing & Mathematical Sciences（计算与数学科学） ； Computation & Neural Systems（计算与神经系统） ； California Institute of Technology（加州理工学院） 纠错

Comments 10 pages, 5 figures, NeurIPS 2025 Foundation Models for the Brain and Body

The Good, The Bad, and The Hybrid: A Reward Structure Showdown in Reasoning Models Training

Subramanyam Sahoo

机构 * Berkeley AI Safety Initiative (BASIS) UC Berkeley（伯克利人工智能安全计划（BASIS）加州大学伯克利分校） 纠错

Comments Paper accepted to the 2nd Workshop on Aligning Reinforcement Learning Experimentalists and Theorists (ARLET 2025) at NeurIPS; the paper consists of 14 pages (including the appendix) and contains 3 figures

Video Finetuning Improves Reasoning Between Frames

Ruiqi Yang, Tian Yun, Zihan Wang, Ellie Pavlick

机构 * Brown University（布朗大学） 纠错

Comments Accepted at CogInterp @ NeurIPS 2025

Adaptively Coordinating with Novel Partners via Learned Latent Strategies

Benjamin Li, Shuyang Shi, Lucia Romero, Huao Li, Yaqi Xie, Woojun Kim, Stefanos Nikolaidis, Michael Lewis, Katia Sycara, Simon Stepputtis

机构 * Carnegie Mellon University（卡内基梅隆大学） ； University of Pittsburgh（匹兹堡大学） ； University of Southern California（南加州大学） ； Virginia Tech（弗吉尼亚理工学院） 纠错

Comments Accepted to NeurIPS 2025

From Images to Physics: Probabilistic Inference of Galaxy Parameters and Emission Lines via VAE & Normalizing Flows

Adiba Amira Siddiqa, Sayed Shafaat Mahmud, Rafael Martinez-Galarza

Comments 9 pages, 5 figures, Accepted at NeurIPS 2025 Machine Learning and the Physical Sciences Workshop

Toward Real-world Text Image Forgery Localization: Structured and Interpretable Data Synthesis

Zeqin Yu, Haotao Xie, Jian Zhang, Jiangqun Ni, Wenkan Su, Jiwu Huang

机构 * Sun Yat-sen University（中山大学） ； Beihang University（北航） ； Guangzhou University（广州大学） ； Peng Cheng Laboratory（鹏城实验室） ； Shenzhen MSU-BIT University（深圳MSU-BIT大学） 纠错

Comments NeurIPS 2025 D&B Track

PID-controlled Langevin Dynamics for Faster Sampling of Generative Models

Hongyi Chen, Jianhai Shu, Jingtao Ding, Yong Li, Xiao-Ping Zhang

机构 * Shenzhen Key Laboratory of Ubiquitous Data Enabling Laboratory, Shenzhen International Graduate School, Tsinghua University, Shenzhen 518055, China（深圳 ubiquitous 数据赋能实验室，深圳国际研究生院，清华大学，深圳） ； Department of Electronic Engineering, Tsinghua University, Beijing 100084, China（电子工程系，清华大学，北京） ； Pengcheng Laboratory, Shenzhen 518055, China（鹏城实验室，深圳） 纠错

Comments NeurIPS 2025 poster paper

Seg-VAR: Image Segmentation with Visual Autoregressive Modeling

Rongkun Zheng, Lu Qi, Xi Chen, Yi Wang, Kun Wang, Hengshuang Zhao

机构 * The University of Hong Kong（香港大学） ； Insta360 ； Shanghai Artificial Intelligence Laboratory（上海人工智能实验室） ； Shanghai Innovation Institute（上海创新研究院） ； SenseTime Research（商汤科技研究院） 纠错

Comments NeurIPS 2025, 22 pages

Knowledge is Overrated: A zero-knowledge machine learning and cryptographic hashing-based framework for verifiable, low latency inference at the LHC

Pratik Jawahar, Caterina Doglioni, Maurizio Pierini

机构 * University of Manchester（曼彻斯特大学） ； CERN（欧洲核子研究中心） 纠错

Comments ML4PS NeurIPS 2025

LOBERT: Generative AI Foundation Model for Limit Order Book Messages

Eljas Linna, Kestutis Baltakys, Alexandros Iosifidis, Juho Kanniainen

机构 * Data Science Research Centre（数据科学研究中心） ； Tampere University（塔尔库大学） 纠错

Comments Submission for NeurIPS 2025 GenAI in Finance Workshop

Cross-Layer Design for Near-Field mmWave Beam Management and Scheduling under Delay-Sensitive Traffic

Zijun Wang, Anjali Omer, Jacob Chakareski, Nicholas Mastronarde, Rui Zhang

Comments Workshop paper submitted to the AI4NextG Workshop at NeurIPS 2025

Edge Machine Learning for Cluster Counting in Next-Generation Drift Chambers

Deniz Yilmaz, Liangyu Wu, Julia Gonski, Dylan Rankin, Christian Herwig

机构 * Department of Physics（物理系） ； Stanford University（斯坦福大学） ； SLAC National Accelerator Laboratory（SLAC国家加速器实验室） ； University of Pennsylvania（宾夕法尼亚大学） ； University of Michigan（密歇根大学） 纠错

Comments 6 pages, 3 figures, 1 table. Machine Learning and the Physical Sciences Workshop, NeurIPS 2025

Uncovering Strategic Egoism Behaviors in Large Language Models

Yaoyuan Zhang, Aishan Liu, Zonghao Ying, Xianglong Liu, Jiangfan Liu, Yisong Xiao, Qihang Zhang

Comments PersonaNLP@NeurIPS 2025

CDFlow: Building Invertible Layers with Circulant and Diagonal Matrices

Xuchen Feng, Siyu Liao

机构 * School of Integrated Circuits（集成电路学院） ； Sun Yat-sen University（中山大学） 纠错

Comments Accepted at NeurIPS 2025. 10 pages, 12 figures, 2 tables

Online Mixture of Experts: No-Regret Learning for Optimal Collective Decision-Making

Larkin Liu, Jalal Etesami

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

Timely Clinical Diagnosis through Active Test Selection

Silas Ruhrberg Estévez, Nicolás Astorga, Mihaela van der Schaar

Journal ref Proceedings of the 39th International Conference on Neural Information Processing Systems (NeurIPS 2025)

DEXTER: Diffusion-Guided EXplanations with TExtual Reasoning for Vision Models

Simone Carnemolla, Matteo Pennisi, Sarinda Samarasinghe, Giovanni Bellitto, Simone Palazzo, Daniela Giordano, Mubarak Shah, Concetto Spampinato

机构 * University of Catania（卡塔尼亚大学） ； University of Central Florida（中央佛罗里达大学） 纠错

Comments Accepted to NeurIPS 2025 (spotlight)

A Human Behavioral Baseline for Collective Governance in Software Projects

Mobina Noori, Mahasweta Chakraborti, Amy X Zhang, Seth Frey

机构 * Department of Computer Science University of California Davis（计算机科学系加州大学戴维斯分校） ； Department of Communication University of California, Davis（沟通系加州大学戴维斯分校） ； Allen School of Computer Science & Engineering University of Washington（计算机科学与工程学院华盛顿大学） 纠错

Comments Algorithmic Collective Action Workshop @ NeurIPS 2025. arXiv admin note: text overlap with arXiv:2509.16295

InsFusion: Rethink Instance-level LiDAR-Camera Fusion for 3D Object Detection

Zhongyu Xia, Hansong Yang, Yongtao Wang

机构 * Wangxuan Institute of Computer Technology, Peking University（北京大学计算机技术研究院） 纠错

Comments NeurIPS 2025 workshop

NoLBERT: A No Lookahead(back) Foundational Language Model

Ali Kakhbod, Peiyao Li

机构 * Haas School of Business University of California, Berkeley（哈斯商学院 加州大学伯克利分校） 纠错

Journal ref NeurIPS 2025 (GenAI in Finance)

SANSA: Unleashing the Hidden Semantics in SAM2 for Few-Shot Segmentation

Claudia Cuttano, Gabriele Trivigno, Giuseppe Averta, Carlo Masone

机构 * Politecnico di Torino（托里诺理工学院） 纠错

Comments Accepted to NeurIPS 2025 as Spotlight