NeurIPS

Mapping Faithful Reasoning in Language Models

Jiazheng Li, Andreas Damianou, J Rosser, José Luis Redondo García, Konstantina Palla

机构 * King’s College London（伦敦国王学院） ； Spotify UK（Spotify英国分公司） ； University of Oxford（牛津大学） ； Spotify Spain（Spotify西班牙分公司） 纠错

Comments 9 pages, Accepted to the Mechanistic Interpretability Workshop at NeurIPS 2025

Transformer Key-Value Memories Are Nearly as Interpretable as Sparse Autoencoders

Mengyu Ye, Jun Suzuki, Tatsuro Inaba, Tatsuki Kuribayashi

机构 * Tohoku University（东大大学） ； RIKEN（理化学研究所） ； MBZUAI（马克斯·普朗克人工智能研究所） 纠错

Comments NeurIPS 2025

Accident Anticipation via Temporal Occurrence Prediction

Tianhao Zhao, Yiyang Zou, Zihao Mao, Peilun Xiao, Yulin Huang, Hongda Yang, Yuxuan Li, Qun Li, Guobin Wu, Yutian Lin

机构 * School of Computer Science, Wuhan University（武汉大学计算机学院） ； Zhongguancun Academy（中关村学院） ； Didi Chuxing（滴滴出行） 纠错

Comments Accepted by NIPS 2025

LUNA: Efficient and Topology-Agnostic Foundation Model for EEG Signal Analysis

Berkay Döner, Thorir Mar Ingolfsson, Luca Benini, Yawei Li

机构 * Integrated Systems Laboratory, ETH Zürich（瑞士苏黎世联邦理工学院集成系统实验室） 纠错

Comments NeurIPS camera-ready version, 27 pages, 10 figures, 13 tables

Diffusion-Driven Two-Stage Active Learning for Low-Budget Semantic Segmentation

Jeongin Kim, Wonho Bae, YouLee Han, Giyeong Oh, Youngjae Yu, Danica J. Sutherland, Junhyug Noh

Comments Accepted to NeurIPS 2025

OptiTree: Hierarchical Thoughts Generation with Tree Search for LLM Optimization Modeling

Haoyang Liu, Jie Wang, Yuyang Cai, Xiongwei Han, Yufei Kuang, Jianye Hao

机构 * MoE Key Laboratory of Brain-inspired Intelligent Perception and Cognition, University of Science and Technology of China（脑启发式感知与认知联合实验室，中国科学技术大学） ； Noah’s Ark Lab, Huawei Technologies（华为诺亚实验室） ； Tianjin University（天津大学） 纠错

Comments Published at NeurIPS 2025

Towards General Modality Translation with Contrastive and Predictive Latent Diffusion Bridge

Nimrod Berman, Omkar Joglekar, Eitan Kosman, Dotan Di Castro, Omri Azencot

机构 * Bosch AI Center（博世人工智能中心） ； Ben-Gurion University of the Negev（巴伊兰大学） ； Technical University of Munich（慕尼黑技术大学） 纠错

Comments Accepted as a poster at NeurIPS 2025

Beyond Fertility: Analyzing STRR as a Metric for Multilingual Tokenization Evaluation

Mir Tafseer Nayeem, Sawsan Alqahtani, Md Tahmid Rahman Laskar, Tasnim Mohiuddin, M Saiful Bari

机构 * University of Alberta（阿尔伯塔大学） ； Princess Nourah Bint Abdulrahman University（努拉·阿卜杜勒拉赫曼王后大学） ； Dialpad Qatar Computing Research Institute（阿联酋迪拜计算研究所） ； Amazon AGI（亚马逊人工智能研究所） 纠错

Comments NeurIPS 2025 Workshop

One Stone with Two Birds: A Null-Text-Null Frequency-Aware Diffusion Models for Text-Guided Image Inpainting

Haipeng Liu, Yang Wang, Meng Wang

Comments 27 pages, 11 figures, to appear at NeurIPS 2025

Retrv-R1: A Reasoning-Driven MLLM Framework for Universal and Efficient Multimodal Retrieval

Lanyun Zhu, Deyi Ji, Tianrun Chen, Haiyang Wu, Shiqi Wang

机构 * City University of Hong Kong（香港城市大学） ； Tencent（腾讯） ； Zhejiang University（浙江大学） 纠错

Comments NeurIPS 2025

EuroSpeech: A Multilingual Speech Corpus

Samuel Pfisterer, Florian Grötschla, Luca A. Lanzendörfer, Florian Yan, Roger Wattenhofer

机构 * ETH Zurich（苏黎世联邦理工学院） 纠错

Comments Published in the 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Track on Datasets and Benchmark

Differentiable Structure Learning and Causal Discovery for General Binary Data

Chang Deng, Bryon Aragam

Comments 30 pages, 6 figures, to appear at NeurIPS 2025

SceneWeaver: All-in-One 3D Scene Synthesis with an Extensible and Self-Reflective Agent

Yandan Yang, Baoxiong Jia, Shujie Zhang, Siyuan Huang

机构 * State Key Laboratory of General Artificial Intelligence, BIGAI（1 通用人工智能国家重点实验室、BIGAI） ； Tsinghua University（2 清华大学） 纠错

Comments Accepted by NeurIPS 2025, 26 pages

Self-Supervised Discovery of Neural Circuits in Spatially Patterned Neural Responses with Graph Neural Networks

Kijung Yoon

机构 * Department of Electronic Engineering（电子工程系） ； Department of Artificial Intelligence（人工智能系） ； Hanyang University（翰阳大学） 纠错

Comments To appear in NeurIPS 2025

EvoBrain: Dynamic Multi-Channel EEG Graph Modeling for Time-Evolving Brain Networks

Rikuto Kotoge, Zheng Chen, Tasuku Kimura, Yasuko Matsubara, Takufumi Yanagisawa, Haruhiko Kishima, Yasushi Sakurai

机构 * Department of Neurosurgery, Graduate School of Medicine, The University of Osaka, Japan（大阪大学医学研究生院神经外科） ； Institute for Advanced Co-Creation Studies, The University of Osaka, Japan（大阪大学先进协同研究机构） 纠错

Comments Accepted by NeurIPS 2025 (spotlight)

Beyond the Average: Distributional Causal Inference under Imperfect Compliance

Undral Byambadalai, Tomu Hirata, Tatsushi Oka, Shota Yasui

Comments Proceedings of the Neural Information Processing Systems (NeurIPS) 2025

BTL-UI: Blink-Think-Link Reasoning Model for GUI Agent

Shaojie Zhang, Ruoceng Zhang, Pei Fu, Shaokang Wang, Jiahui Yang, Xin Du, Shiqi Cui, Bin Qin, Ying Huang, Zhenbo Luo, Jian Luan

机构 * Xiaomi Inc（小米公司） 纠错

Comments Accepted at NeurIPS 2025

Score-informed Neural Operator for Enhancing Ordering-based Causal Discovery

Jiyeon Kang, Songseong Kim, Chanhui Lee, Doyeong Hwang, Joanie Hayoun Chung, Yunkyung Ko, Sumin Lee, Sungwoong Kim, Sungbin Lim

Comments Accepted to NeurIPS 2025. 36 pages, 18 figures, 12 tables

Ground-Compose-Reinforce: Grounding Language in Agentic Behaviours using Limited Data

Andrew C. Li, Toryn Q. Klassen, Andrew Wang, Parand A. Alamdari, Sheila A. McIlraith

机构 * Department of Computer Science, University of Toronto（计算机科学系，多伦多大学） ； Vector Institute for Artificial Intelligence（人工智能矢量研究所） ； Schwartz Reisman Institute for Technology and Society（技术与社会施瓦茨-雷曼研究所） 纠错

Comments NeurIPS 2025

DATE-LM: Benchmarking Data Attribution Evaluation for Large Language Models

Cathy Jiao, Yijun Pan, Emily Xiao, Daisy Sheng, Niket Jain, Hanzhang Zhao, Ishita Dasgupta, Jiaqi W. Ma, Chenyan Xiong

机构 * Carnegie Mellon University（卡内基梅隆大学） ； University of Michigan（密歇根大学） ； UIUC（伊利诺伊大学香槟分校） 纠错

Comments NeurIPS 2025 Datasets and Benchmarks Track

Spurious-Aware Prototype Refinement for Reliable Out-of-Distribution Detection

Reihaneh Zohrabi, Hosein Hasani, Mahdieh Soleymani Baghshah, Anna Rohrbach, Marcus Rohrbach, Mohammad Hossein Rohban

机构 * TU Darmstadt（图宾根大学） ； Sharif University of Technology（谢赫·穆罕默德·卡齐姆大学） 纠错

Comments Accepted at NeurIPS 2025

Complexity Scaling Laws for Neural Models using Combinatorial Optimization

Lowell Weissman, Michael Krumdick, A. Lynn Abbott

机构 * Virginia Tech（弗吉尼亚理工大学） ； Kensho Technologies（Kensho科技） 纠错

Comments NeurIPS 2025, 46 pages, 20 figures

Variational Learning Finds Flatter Solutions at the Edge of Stability

Avrajit Ghosh, Bai Cong, Rio Yokota, Saiprasad Ravishankar, Rongrong Wang, Molei Tao, Mohammad Emtiyaz Khan, Thomas Möllenhoff

机构 * Michigan State University（密歇根州立大学） ； Institute of Science Tokyo（东京科学研究所） ； RIKEN Center for AI Project（理化学研究所人工智能项目中心） ； Georgia Institute of Technology（佐治亚理工学院） 纠错

Journal ref NeurIPS-2025

3D-RAD: A Comprehensive 3D Radiology Med-VQA Dataset with Multi-Temporal Analysis and Diverse Diagnostic Tasks

Xiaotang Gai, Jiaxiang Liu, Yichen Li, Zijie Meng, Jian Wu, Zuozhu Liu

机构 * ZJU-Angelalign R&D Center for Intelligence Healthcare（浙江大学智能医疗研发中心） ； Zhejiang University（浙江大学） ； Zhejiang Key Laboratory of Medical Imaging Artificial Intelligence（浙江省医学影像人工智能重点实验室） ； Guangdong Institute of Intelligence Science and Technology（广东省智能科学与技术研究院） 纠错

Comments Accepted by NeurIPS 2025

Are Pixel-Wise Metrics Reliable for Sparse-View Computed Tomography Reconstruction?

Tianyu Lin, Xinran Li, Chuntung Zhuang, Qi Chen, Yuanhao Cai, Kai Ding, Alan L. Yuille, Zongwei Zhou

机构 * Johns Hopkins University（约翰霍普金斯大学） ； Yale University（耶鲁大学） ； Johns Hopkins Medicine（约翰霍普金斯医学） 纠错

Comments NeurIPS 2025

SUMO: Subspace-Aware Moment-Orthogonalization for Accelerating Memory-Efficient LLM Training

Yehonathan Refael, Guy Smorodinsky, Tom Tirer, Ofir Lindenbaum

机构 * Faculty of Engineering（工程学院） ； Tel Aviv University（特拉维夫大学） ； Department of Computer science（计算机科学系） ； Ben Gurion University（本· Gurion大学） ； Faculty of Engineering Bar-Ilan University（巴伊兰大学工程学院） ； Bar-Ilan University（巴伊兰大学） 纠错

Journal ref The Thirty-Ninth Annual Conference on Neural Information Processing Systems (NeurIPS 2025)

A Simple Linear Patch Revives Layer-Pruned Large Language Models

Xinrui Chen, Haoli Bai, Tao Yuan, Ruikang Liu, Kang Zhao, Xianzhi Yu, Lu Hou, Tian Guan, Yonghong He, Chun Yuan

机构 * Shenzhen International Graduate School, Tsinghua University（清华大学深圳国际研究生院） ； Huawei Technologies（华为技术） 纠错

Comments 26 pages, accepted to NeurIPS 2025

On the Surprising Effectiveness of Large Learning Rates under Standard Width Scaling

Moritz Haas, Sebastian Bordt, Ulrike von Luxburg, Leena Chennuru Vankadara

机构 * University of Tübingen（图宾根大学） ； Tübingen AI Center（图宾根人工智能中心） ； Gatsby Computational Neuroscience Unit（Gatsby计算神经科学单位） ； University College London（伦敦大学学院） 纠错

Comments NeurIPS 2025 (spotlight) camera-ready version. Open source code for reproducing our experiments can be found under https://github.com/moritzhaas/large-lr-width-scaling Open and easily adaptable code that implements fine-grained tracking of neural network internal statistics can be found under https://github.com/tml-tuebingen/torch-module-monitor

First SFT, Second RL, Third UPT: Continual Improving Multi-Modal LLM Reasoning via Unsupervised Post-Training

Lai Wei, Yuting Li, Chen Wang, Yue Wang, Linghe Kong, Weiran Huang, Lichao Sun

机构 * School of Computer Science, Shanghai Jiao Tong University（上海交通大学计算机科学学院） ； Zhongguancun Academy（中关村学院） ； Shanghai Innovation Institute（上海创新研究院） ； Lehigh University（莱特大学） 纠错

Comments Accepted by NeurIPS 2025

OVERT: A Benchmark for Over-Refusal Evaluation on Text-to-Image Models

Ziheng Cheng, Yixiao Huang, Hui Xu, Somayeh Sojoudi, Xuandong Zhao, Dawn Song, Song Mei

机构 * UC Berkeley（加州大学伯克利分校） 纠错

Comments NeurIPS 2025 (D&B Track)