NeurIPS - arXivDaily 期刊&会议

2509.21310 2025-09-26 cs.AI

SAGE: A Realistic Benchmark for Semantic Understanding

Samarth Goel, Reagan J. Lee, Kannan Ramchandran

机构 * University of California, Berkeley（加州大学伯克利分校）

Comments 39th Conference on Neural Information Processing Systems (NeurIPS 2025) Workshop: Evaluating the Evolving LLM Lifecycle: Benchmarks, Emergent Abilities, and Scaling

2509.21254 2025-09-26 cs.LG math.OC

humancompatible.train: Implementing Optimization Algorithms for Stochastically-Constrained Stochastic Optimization Problems

Andrii Kliachkin, Jana Lepšová, Gilles Bareilles, Jakub Mareček

机构 * Artificial Intelligence Center, Czech Technical University in Prague（捷克技术大学布拉格分校人工智能中心）

Comments Accepted at NeurIPS workshop COML 2025

2509.21247 2025-09-26 cs.CV cs.AI

Learning to Look: Cognitive Attention Alignment with Vision-Language Models

Ryan L. Yang, Dipkamal Bhusal, Nidhi Rastogi

机构 * Brown University（布朗大学）； Rochester Institute of Technology（罗切斯特理工大学）

Comments 7 pages, neurips workshop

2509.20998 2025-09-26 cs.AI

CORE: Full-Path Evaluation of LLM Agents Beyond Final State

Panagiotis Michelakis, Yiannis Hadjiyiannis, Dimitrios Stamoulis

机构 * Synkrasis Labs（Synkrasis实验室）； Harbin Institute of Technology（哈尔滨工业大学）

Comments Accepted: LAW 2025 Workshop NeurIPS 2025

2509.20891 2025-09-26 cs.SD

AIBA: Attention-based Instrument Band Alignment for Text-to-Audio Diffusion

Junyoung Koh, Soo Yong Kim, Gyu Hyeong Choi, Yongwon Choi

机构 * Department of Artificial Intelligence, Yonsei University（人工智能系，延世大学）； MAAP LAB, MODULABS（MODULABS 音频实验室）； KRAFTON ； AI Matics ； Department of Media Software, Sungkyul University（媒体软件系，松谷大学）

Comments NeurIPS 2025 AI for Music Workshop

2509.20824 2025-09-26 cs.GR cs.CV

ARMesh: Autoregressive Mesh Generation via Next-Level-of-Detail Prediction

Jiabao Lei, Kewei Shi, Zhihao Liang, Kui Jia

机构 * The Chinese University of Hong Kong, Shenzhen（香港中文大学（深圳））； The University of Hong Kong（香港大学）； Tencent Hunyuan（腾讯文心）

Comments NeurIPS 2025, Project Page: https://jblei.site/proj/armesh

2509.20683 2025-09-26 physics.flu-dyn cs.LG

Implicit Augmentation from Distributional Symmetry in Turbulence Super-Resolution

Julia Balla, Jeremiah Bailey, Ali Backour, Elyssa Hofgard, Tommi Jaakkola, Tess Smidt, Ryley McConkey

机构 * Massachusetts Institute of Technology（麻省理工学院）

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

2509.20678 2025-09-26 cs.LG cs.AI cs.CV stat.ML

Bispectral OT: Dataset Comparison using Symmetry-Aware Optimal Transport

Annabel Ma, Kaiying Hou, David Alvarez-Melis, Melanie Weber

机构 * Harvard University（哈佛大学）

Comments Accepted to NeurIPS 2025 Workshop on Symmetry and Geometry in Neural Representations (NeurReps)

2509.20570 2025-09-26 cs.LG cs.AI cs.CE cs.SY eess.SY

PIRF: Physics-Informed Reward Fine-Tuning for Diffusion Models

Mingze Yuan, Pengfei Jin, Na Li, Quanzheng Li

机构 * Harvard University（哈佛大学）； Massachusetts General Hospital（麻省总医院）

Comments 18 pages, 6 figures; NeurIPS 2025 AI for science workshop

2509.20549 2025-09-26 cs.LG cs.AI

Understanding and Improving Adversarial Robustness of Neural Probabilistic Circuits

Weixin Chen, Han Zhao

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

Comments NeurIPS 2025 Camera Ready

2509.20461 2025-09-26 cs.CL cs.LG

Document Summarization with Conformal Importance Guarantees

Bruce Kuwahara, Chen-Yuan Lin, Xiao Shi Huang, Kin Kwan Leung, Jullian Arta Yapeter, Ilya Stanevich, Felipe Perez, Jesse C. Cresswell

机构 * Signal 1 AI ； Layer 6 AI

Comments NeurIPS 2025. Code is available at https://github.com/layer6ai-labs/conformal-importance-summarization

2509.20408 2025-09-26 cs.LG cs.DC

A Theory of Multi-Agent Generative Flow Networks

Leo Maxime Brunswic, Haozhi Wang, Shuang Luo, Jianye Hao, Amir Rasouli, Yinchuan Li

机构 * Huawei Technologies Canada（华为加拿大公司）； Tianjin University（天津大学）； Zhejiang University（浙江大学）； Huawei Noah’s Ark Lab（华为诺亚实验室）

Comments Accepted at SPIGM Workshop NeurIPS 2025

2509.20383 2025-09-26 cs.CR cs.AI

MARS: A Malignity-Aware Backdoor Defense in Federated Learning

Wei Wan, Yuxuan Ning, Zhicong Huang, Cheng Hong, Shengshan Hu, Ziqi Zhou, Yechao Zhang, Tianqing Zhu, Wanlei Zhou, Leo Yu Zhang

机构 * Faculty of Data Science, City University of Macau（澳门城市大学数据科学学院）； School of Computing, Australian National University（澳大利亚国立大学计算机学院）； Ant Group（蚂蚁集团）； School of Cyber Science and Engineering, Huazhong University of Science and Technology（华中科技大学网络安全工程学院）； School of Computer Science and Technology, Huazhong University of Science and Technology（华中科技大学计算机科学与技术学院）； College of Computing and Data Science, Nanyang Technological University（南洋理工大学计算与数据科学学院）； School of Information and Communication Technology, Griffith University（格里菲斯大学信息与通信技术学院）

Comments NeurIPS 2025

2509.18056 2025-09-26 cs.CV

TempSamp-R1: Effective Temporal Sampling with Reinforcement Fine-Tuning for Video LLMs

Yunheng Li, Jing Cheng, Shaoyong Jia, Hangyi Kuang, Shaohui Jiao, Qibin Hou, Ming-Ming Cheng

机构 * VCIP, School of Computer Science, Nankai University（VCIP，计算机科学学院，南开大学）； ByteDance Inc.（字节跳动公司）

Comments Accepted at NeurIPS 2025

2509.16421 2025-09-26 cs.CV cs.AI

AHA -- Predicting What Matters Next: Online Highlight Detection Without Looking Ahead

Aiden Chang, Celso De Melo, Stephanie M. Lukin

机构 * University of Southern California（南加州大学）； DEVCOM Army Research Laboratory（陆军研究实验室）

Comments Accepted at NeurIPS 2025, 32 pages, 5 figures

2508.07807 2025-09-26 cs.LG

Topological Feature Compression for Molecular Graph Neural Networks

Rahul Khorana

Comments Accepted at Thirty-ninth Conference on Neural Information Processing Systems (NeurIPS) 2025 AI4Science Workshop

Journal ref Thirty-ninth Conference on Neural Information Processing Systems (NeurIPS) 2025 AI4Science Workshop

2506.18896 2025-09-26 cs.CL

ReasonFlux-PRM: Trajectory-Aware PRMs for Long Chain-of-Thought Reasoning in LLMs

Jiaru Zou, Ling Yang, Jingwen Gu, Jiahao Qiu, Ke Shen, Jingrui He, Mengdi Wang

机构 * UIUC（伊利诺伊大学香槟分校）； Princeton University（普林斯顿大学）； Cornell University（康奈尔大学）； ByteDance（字节跳动）

Comments Accepted by NeurIPS 2025. Project: https://github.com/Gen-Verse/ReasonFlux

2506.03136 2025-09-26 cs.CL

Co-Evolving LLM Coder and Unit Tester via Reinforcement Learning

Yinjie Wang, Ling Yang, Ye Tian, Ke Shen, Mengdi Wang

机构 * University of Chicago（芝加哥大学）； Princeton University（普林斯顿大学）； Peking University（北京大学）； ByteDance Seed（字节跳动种子）

Comments NeurIPS 2025 Spotlight. Project: https://github.com/Gen-Verse/CURE

2505.15809 2025-09-26 cs.CV

MMaDA: Multimodal Large Diffusion Language Models

Ling Yang, Ye Tian, Bowen Li, Xinchen Zhang, Ke Shen, Yunhai Tong, Mengdi Wang

机构 * Princeton University（普林斯顿大学）； Peking University（北京大学）； Tsinghua University（清华大学）； ByteDance Seed（字节跳动种子）

Comments NeurIPS 2025. Project: https://github.com/Gen-Verse/MMaDA

2505.15155 2025-09-26 q-fin.CP cs.AI cs.CE cs.LG

R&D-Agent-Quant: A Multi-Agent Framework for Data-Centric Factors and Model Joint Optimization

Yuante Li, Xu Yang, Xiao Yang, Minrui Xu, Xisen Wang, Weiqing Liu, Jiang Bian

机构 * Carnegie Mellon University（卡内基梅隆大学）； Microsoft Research Asia（微软亚洲研究院）； Hong Kong University of Science and Technology（香港科学与技术大学）； University of Oxford（牛津大学）

Comments 42 pages,11figures, NeurIPS 2025

2505.10039 2025-09-26 cs.LG

Rethinking Circuit Completeness in Language Models: AND, OR, and ADDER Gates

Hang Chen, Jiaying Zhu, Xinyu Yang, Wenya Wang

机构 * Hang Chen School of Computer Science and Technology Xi’an Jiaotong University（Hang Chen 计算机科学与技术学院西安交通大学）； Jiaying Zhu School of Computer Science and Engineering The Chinese University of Hong Kong（Jiaying Zhu 计算科学与工程学院香港中文大学）； Xinyu Yang School of Computer Science and Technology Xi’an Jiaotong University（Xinyu Yang 计算机科学与技术学院西安交通大学）； Wenya Wang School of Computer Science and Engineering Nanyang Technological University（Wenya Wang 计算科学与工程学院新加坡国立大学）

Comments accepted by NeurIPS 2025 (poster)

2502.12148 2025-09-26 cs.CV

HermesFlow: Seamlessly Closing the Gap in Multimodal Understanding and Generation

Ling Yang, Xinchen Zhang, Ye Tian, Chenming Shang, Minghao Xu, Wentao Zhang, Bin Cui

机构 * Peking University（北京大学）； Tsinghua University（清华大学）； Mila - Québec AI Institute（魁北克人工智能研究所）

Comments NeurIPS 2025. Code: https://github.com/Gen-Verse/HermesFlow

2501.19215 2025-09-26 cs.LG cs.AI

Strassen Attention, Split VC Dimension and Compositionality in Transformers

Alexander Kozachinskiy, Felipe Urrutia, Hector Jimenez, Tomasz Steifer, Germán Pizarro, Matías Fuentes, Francisco Meza, Cristian B. Calderon, Cristóbal Rojas

机构 * University of Chile（智利大学）； CENIA ； IPPT PAN（波兰国家物理与天文学研究所）； IMC UC（UC大学医学中心）

Comments Accepted at NeurIPS 2025

2410.07548 2025-09-26 stat.ML astro-ph.CO cs.IT cs.LG math.IT physics.data-an

Hybrid Summary Statistics

T. Lucas Makinen, Ce Sui, Benjamin D. Wandelt, Natalia Porqueres, Alan Heavens

机构 * Imperial College London（伦敦帝国学院）； Tsinghua University（清华大学）； Sorbonne Université（索邦大学）； Center for Computational Astrophysics, Flatiron Institute（Flatiron研究所计算天体物理学中心）； Oxford University（牛津大学）

Comments 7 pages, 4 figures. Accepted to ML4PS2024 at NeurIPS 2024. Code available at https://github.com/tlmakinen/hybridStats

2402.06353 2025-09-26 cs.CV cs.CY cs.DL cs.LG

Copycats: the many lives of a publicly available medical imaging dataset

Amelia Jiménez-Sánchez, Natalia-Rozalia Avlona, Dovile Juodelyte, Théo Sourget, Caroline Vang-Larsen, Anna Rogers, Hubert Dariusz Zając, Veronika Cheplygina

机构 * IT University of Copenhagen（哥本哈根IT大学）； University of Copenhagen（哥本哈根大学）

Comments NeurIPS 2024 Track on Datasets and Benchmarks. Please note that v1 has a different title