ICML

Cover Learning for Large-Scale Topology Representation

Luis Scoccola, Uzu Lim, Heather A. Harrington

机构 * Centre de Recherches Mathématiques et Institut des sciences mathématiques, Laboratoire de combinatoire et d'informatique mathématique de l'Université du Québec à Montréal, Université de Sherbrooke, Canada（数学研究中心和数学科学研究所，魁北克大学蒙特利尔分校组合与数学信息学实验室， Sherbrooke 大学，加拿大） ； Queen Mary University of London, United Kingdom（伦敦女王学院，英国） ； Max Planck Institute of Molecular Cell Biology and Genetics（分子细胞生物学与遗传学马克斯·普朗克研究所） ； Centre for Systems Biology, Dresden, Germany（系统生物学中心，德累斯顿，德国） ； Faculty of Mathematics, Technische Universität Dresden, Germany（数学系，德累斯顿技术大学，德国） ； Mathematical Institute, University of Oxford, United Kingdom（数学研究所，牛津大学，英国） 纠错

Comments 29 pages, 19 figures, 5 tables; final version at ICML 2025

LVADNet3D: A Deep Autoencoder for Reconstructing 3D Intraventricular Flow from Sparse Hemodynamic Data

Mohammad Abdul Hafeez Khan, Marcello Mattei Di Eugeni, Benjamin Diaz, Ruth E. White, Siddhartha Bhattacharyya, Venkat Keshav Chivukula

机构 * Florida Institute of Technology, USA（佛罗里达理工学院） 纠错

Comments Accepted to International Conference on Machine Learning and Applications (ICMLA), 6 pages, 4 figure, 3 tables

Code Generation with Small Language Models: A Codeforces-Based Study

Débora Souza, Rohit Gheyi, Lucas Albuquerque, Gustavo Soares, Márcio Ribeiro

Comments Accepted at International Conference on Machine Learning and Applications (ICMLA 2025)

WikiBigEdit: Understanding the Limits of Lifelong Knowledge Editing in LLMs

Lukas Thede, Karsten Roth, Matthias Bethge, Zeynep Akata, Tom Hartvigsen

机构 * Tübingen AI Center, University of Tübingen（图宾根人工智能中心，图宾根大学） ； Munich Center for Machine Learning (MCML)（慕尼黑机器学习中心） ； Technical University, Munich（慕尼黑技术大学） ； University of Virginia（弗吉尼亚大学） 纠错

Comments published at ICML 2025

Controlling Neural Collapse Enhances Out-of-Distribution Detection and Transfer Learning

Md Yousuf Harun, Jhair Gallardo, Christopher Kanan

机构 * Rochester Institute of Technology（罗切斯特技术学院） ； University of Rochester（罗切斯特大学） 纠错

Comments Published in ICML 2025

Mental Multi-class Classification on Social Media: Benchmarking Transformer Architectures against LSTM Models

Khalid Hasan, Jamil Saquer, Yifan Zhang

机构 * Dept. of Computer Science Missouri State University Springfield, MO, USA（计算机科学系 莫斯科州立大学 斯普林菲尔德 MO 美国） 纠错

Comments 24th IEEE International Conference on Machine Learning and Applications, ICMLA 2025 (camera-ready)

Journal ref 2025 International Conference on Machine Learning and Applications (ICMLA)

Entropic Causal Inference: Graph Identifiability

Spencer Compton, Kristjan Greenewald, Dmitriy Katz, Murat Kocaoglu

机构 * Massachusetts Institute of Technology（麻省理工学院） ； MIT-IBM Watson AI Lab（MIT-IBM Watson AI实验室） ； IBM Research（IBM研究院） ； Purdue University（普渡大学） 纠错

Comments Presented at ICML 2022. This version corrects a bug in semi-synthetic experiments

Activation Space Interventions Can Be Transferred Between Large Language Models

Narmeen Oozeer, Dhruv Nathawani, Nirmalendu Prakash, Michael Lan, Abir Harrasse, Amirali Abdullah

机构 * Nvidia ； Singapore University of Technology and Design（新加坡技术与设计大学） 纠错

Comments 75 pages. Accepted to ICML 2025

TICA-Based Free Energy Matching for Machine-Learned Molecular Dynamics

Alexander Aghili, Andy Bruce, Daniel Sabo, Razvan Marinescu

机构 * Baskin Engineering, University of California - Santa Cruz, Santa Cruz, United States（加州大学圣克鲁兹分校贝斯基工程） 纠错

Comments Proceedings of the ICML 2025 Workshop on Multi-modal Foundation Models and Large Language Models for Life Sciences, Vancouver, Canada. 2025. Copyright 2025 by the author(s). 4 Pages 5 Figures

Domain Adaptation for Ulcerative Colitis Severity Estimation Using Patient-Level Diagnoses

Takamasa Yamaguchi, Brian Kenji Iwana, Ryoma Bise, Shota Harada, Takumi Okuo, Kiyohito Tanaka, Kaito Shiku

机构 * Kyushu University, Japan（九州大学） ； Kyoto Second Red Cross Hospital, Japan（京都第二红十字医院） 纠错

Comments Accepted to MICCAI workshop 2025 (International conference on machine learning in medical imaging)

Towards Rationale-Answer Alignment of LVLMs via Self-Rationale Calibration

Yuanchen Wu, Ke Yan, Shouhong Ding, Ziyin Zhou, Xiaoqiang Li

机构 * School of Computer Engineering（计算机工程学院） ； Key Laboratory of Multimedia Trusted Perception（多媒体可信感知关键实验室） ； Efficient Computing, Xiamen University（高效计算，厦门大学） ； Tencent Youtu Lab（腾讯优图实验室） 纠错

Comments Accepted by ICML 2025

AQUA-LLM: Evaluating Accuracy, Quantization, and Adversarial Robustness Trade-offs in LLMs for Cybersecurity Question Answering

Onat Gungor, Roshan Sood, Harold Wang, Tajana Rosing

Comments Accepted by the 24th IEEE International Conference on Machine Learning and Applications (ICMLA'25)

LIGHT-HIDS: A Lightweight and Effective Machine Learning-Based Framework for Robust Host Intrusion Detection

Onat Gungor, Ishaan Kale, Jiasheng Zhou, Tajana Rosing

Comments Accepted by The 24th IEEE International Conference on Machine Learning and Applications (ICMLA'25)

Position: AI Safety Must Embrace an Antifragile Perspective

Ming Jin, Hyunin Lee

Journal ref Proceedings of the 42nd International Conference on Machine Learning, Vancouver, Canada. 2025

Dual Feature Reduction for the Sparse-group Lasso and its Adaptive Variant

Fabio Feser, Marina Evangelou

机构 * Department of Mathematics, Imperial College London, London, UK（帝国理工学院伦敦校区数学系） 纠错

Comments 32 pages, 20 figures, 20 tables

Journal ref International Conference on Machine Learning 42 (2025)

NORA: A Nephrology-Oriented Representation Learning Approach Towards Chronic Kidney Disease Classification

Mohammad Abdul Hafeez Khan, Twisha Bhattacharyya, Omar Khan, Noorah Khan, Alina Aziz Fatima Khan, Mohammed Qutub Khan, Sujoy Ghosh Hajra

机构 * Florida Institute of Technology（佛罗里达理工学院） ； University of California Riverside（加州大学河滨分校） ； Edgewood Jr./Sr. High School（埃德伍德高中） ； Shadan Institute of Medicine Sciences（沙达医学科学研究所） ； Riverside Nephrology Physicians（拉维斯德肾病医生） 纠错

Comments 7 pages, 5 figures, accepted to the International Conference on Machine Learning and Applications (ICMLA) 2025

VADER: A Variational Autoencoder to Infer Planetary Masses and Gas-Dust Disk Properties Around Young Stars

Sayed Shafaat Mahmud, Sayantan Auddy, Neal Turner, Jeffrey S. Bary

机构 * Colgate University（科尔盖特大学） ； Fields Institute（菲尔德研究所） ； Jet Propulsion Laboratory（喷气推进实验室） ； California Institute of Technology（加州理工学院） 纠错

Comments 6 pages, 5 figures, Accepted and Published at International Conference on Machine Learning, Machine Learning for Astrophysics Workshop 2025

TinyServe: Query-Aware Cache Selection for Efficient LLM Serving

Dong Liu, Yanxuan Yu

机构 * Yale University（耶鲁大学） ； Columbia University（哥伦比亚大学） 纠错

Comments Accepted to ACM MM as Oral Paper, also accepted to ICML MOSS workshop, publicly available as https://openreview.net/forum?id=sOdtl4jLci

Test-Time Canonicalization by Foundation Models for Robust Perception

Utkarsh Singhal, Ryan Feng, Stella X. Yu, Atul Prakash

机构 * University of Michigan（密歇根大学） 纠错

Comments Published at ICML 2025

Adversarial Combinatorial Semi-bandits with Graph Feedback

Yuxiao Wen

机构 * Courant Institute of Mathematical Sciences, New York University, New York, USA（Courant数学科学研究所，纽约大学，纽约，美国） 纠错

Comments To appear in ICML 2025

Spec-LLaVA: Accelerating Vision-Language Models with Dynamic Tree-Based Speculative Decoding

Mingxiao Huo, Jiayi Zhang, Hewei Wang, Jinfeng Xu, Zheyu Chen, Huilin Tai, Yijun Chen

机构 * Carnegie Mellon University（卡内基梅隆大学） ； University of Nottingham（诺丁汉大学） ； The University of Hong Kong（香港大学） ； The Hong Kong Polytechnic University（香港理工大学） ； Columbia University（哥伦比亚大学） ； University of California, Berkeley（加州大学伯克利分校） 纠错

Comments 7pages, accepted by ICML TTODLer-FM workshop

MMORE: Massive Multimodal Open RAG & Extraction

Alexandre Sallinen, Stefan Krsteski, Paul Teiletche, Marc-Antoine Allard, Baptiste Lecoeur, Michael Zhang, Fabrice Nemo, David Kalajdzic, Matthias Meyer, Mary-Anne Hartley

机构 * École Polytechnique Fédérale de Lausanne (EPFL), Switzerland（瑞士联邦理工学院洛桑校区） ； ETH Zürich, Switzerland（瑞士苏黎世联邦理工学院） ； T.H. Chan School of Public Health, Harvard University, USA（哈佛大学T.H. Chan公共卫生学院） 纠错

Comments This paper was originally submitted to the CODEML workshop for ICML 2025. 9 pages (including references and appendices)

Integrating Prior Observations for Incremental 3D Scene Graph Prediction

Marian Renz, Felix Igelbrink, Martin Atzmueller

机构 * DFKI Niedersachsen（德克萨斯联合研究所（北莱茵威斯特法伦）） ； Cooperative and Autonomous Systems, DFKI Niedersachsen（合作与自主系统，DFKI北莱茵威斯特法伦） ； German Research Center for Artificial Intelligence（德国人工智能研究中心） ； Semantic Information Systems, Osnabrück University（语义信息系统，奥斯纳布吕克大学） 纠错

Comments Accepted at 24th International Conference on Machine Learning and Applications (ICMLA'25)

DRAG: Data Reconstruction Attack using Guided Diffusion

Wa-Kin Lei, Jun-Cheng Chen, Shang-Tse Chen

机构 * Research Center for Information Technology Innovation, Academia Sinica（中国科学院信息科技创新研究中心） ； National Taiwan University（国立台湾大学） 纠错

Comments ICML 2025

Drug Repurposing Using Deep Embedded Clustering and Graph Neural Networks

Luke Delzer, Robert Kroleski, Ali K. AlShami, Jugal Kalita

机构 * University of Colorado Colorado Springs（科罗拉多州立大学） 纠错

Comments Accepted at the 2025 International Conference on Machine Learning and Applications (ICMLA)

Soft Reasoning: Navigating Solution Spaces in Large Language Models through Controlled Embedding Exploration

Qinglin Zhu, Runcong Zhao, Hanqi Yan, Yulan He, Yudong Chen, Lin Gui

机构 * King's College London, UK（伦敦国王学院） ； The Alan Turing Institute, UK（艾伦·图灵研究所） ； University of Warwick, UK（沃里克大学） 纠错

Comments Accepted as a Spotlight at ICML 2025

Towards LLM Agents for Earth Observation

Chia Hsiang Kao, Wenting Zhao, Shreelekha Revankar, Samuel Speas, Snehal Bhagat, Rajeev Datta, Cheng Perng Phoo, Utkarsh Mall, Carl Vondrick, Kavita Bala, Bharath Hariharan

机构 * Cornell University（康奈尔大学） ； Columbia University（哥伦比亚大学） 纠错

Comments Accepted at ICML 2025 Workshop TerraBytes

LML: A Novel Lexicon for the Moral Foundation of Liberty

Oscar Araque, Lorenzo Gatti, Sergio Consoli, Kyriaki Kalimeri

机构 * Universidad Polit\'ecnica de Madrid, Spain University of Twente, Enschede, The Netherlands European Commission, Joint Research Centre (DG JRC), Ispra, Italy ISI Foundation, Turin, Italy 纠错

Comments Published in the 11th International Conference on Machine Learning, Optimization, and Data Science

RouteFinder: Towards Foundation Models for Vehicle Routing Problems

Federico Berto, Chuanbo Hua, Nayeli Gast Zepeda, André Hottung, Niels Wouda, Leon Lan, Junyoung Park, Kevin Tierney, Jinkyoo Park

机构 * KAIST（韩国科学技术院） ； Bielefeld University（比勒菲尔德大学） ； Rotterdam School of Management（鹿特丹管理学院） ； VU Amsterdam（阿姆斯特丹自由大学） ； Omelet AI4CO 纠错

Comments TMLR 2025. A previous version of this work was presented as Oral at the ICML 2024 FM-Wild Workshop

Steering Protein Language Models

Long-Kai Huang, Rongyi Zhu, Bing He, Jianhua Yao

机构 * Tencent AI Lab（腾讯人工智能实验室） 纠错

Comments Accepted to ICML 2025