ICLR

Sharpness-Aware Data Poisoning Attack

Pengfei He, Han Xu, Jie Ren, Yingqian Cui, Hui Liu, Charu C. Aggarwal, Jiliang Tang

Comments ICLR 2024 Spotlight

Sensitivity-Constrained Fourier Neural Operators for Forward and Inverse Problems in Parametric Differential Equations

Abdolmehdi Behroozi, Chaopeng Shen and, Daniel Kifer

机构 * Department of Civil and Environmental Engineering（土木与环境工程系） ； School of Electrical Engineering and Computer Science（电气与计算机工程学院） 纠错

Comments Accepted and presented at ICLR 2025

Topological Schrödinger Bridge Matching

Maosheng Yang

机构 * Delft University of Technology（代尔夫特理工大学） 纠错

Comments ICLR 2025 Spotlight, 42 pages

RxRx3-core: Benchmarking drug-target interactions in High-Content Microscopy

Oren Kraus, Federico Comitani, John Urbanik, Kian Kenyon-Dean, Lakshmanan Arumugam, Saber Saberian, Cas Wognum, Safiye Celik, Imran S. Haque

机构 * Recursion ； Valence Labs 纠错

Comments Published at LMRL Workshop at ICLR 2025; Updated gene-gene interaction results based on filtering compounds wells out before computing known_relationship_benchmark

GSBA$^K$: $top$-$K$ Geometric Score-based Black-box Attack

Md Farhamdur Reza, Richeng Jin, Tianfu Wu, Huaiyu Dai

机构 * NC State University（北卡罗来纳州立大学） ； Zhejiang University（浙江大学） 纠错

Comments License changed to CC BY 4.0 to align with ICLR 2025. No changes to content. Published at: https://openreview.net/forum?id=htX7AoHyln

Scaling Diffusion Language Models via Adaptation from Autoregressive Models

Shansan Gong, Shivam Agarwal, Yizhe Zhang, Jiacheng Ye, Lin Zheng, Mukai Li, Chenxin An, Peilin Zhao, Wei Bi, Jiawei Han, Hao Peng, Lingpeng Kong

机构 * The University of Hong Kong（香港大学） ； University of Illinois at Urbana-Champaign（伊利诺伊大学厄巴纳-香槟分校） ； Apple（苹果公司） ； Tencent AI Lab（腾讯AI实验室） 纠错

Comments ICLR 2025. (minor updates) Code: https://github.com/HKUNLP/DiffuLLaMA

Weakly-Supervised Affordance Grounding Guided by Part-Level Semantic Priors

Peiran Xu, Yadong Mu

Comments ICLR 2025

Controllable Context Sensitivity and the Knob Behind It

Julian Minder, Kevin Du, Niklas Stoehr, Giovanni Monea, Chris Wendler, Robert West, Ryan Cotterell

机构 * ETH Zürich（苏黎世联邦理工学院） ； EPFL（苏黎世联邦理工学院） ； Cornell University（康奈尔大学） 纠错

Comments Published as a conference paper at ICLR 2025

Agent Security Bench (ASB): Formalizing and Benchmarking Attacks and Defenses in LLM-based Agents

Hanrong Zhang, Jingyuan Huang, Kai Mei, Yifei Yao, Zhenting Wang, Chenlu Zhan, Hongwei Wang, Yongfeng Zhang

机构 * Zhejiang University（浙江大学） ； Rutgers University（罗格斯大学） 纠错

Comments Accepted by ICLR 2025

A Statistical Framework for Ranking LLM-Based Chatbots

Siavash Ameli, Siyuan Zhuang, Ion Stoica, Michael W. Mahoney

机构 * ICSI and Department of Statistics, University of California, Berkeley（ICSI和统计学系，加州大学伯克利分校） ； Department of Computer Science, University of California, Berkeley（计算机科学系，加州大学伯克利分校） ； ICSI, LBNL, and Department of Statistics, University of California, Berkeley（ICSI、劳伦斯伯克利国家实验室和统计学系，加州大学伯克利分校） 纠错

Journal ref The Thirteenth International Conference on Learning Representations (2025)

DaWin: Training-free Dynamic Weight Interpolation for Robust Adaptation

Changdae Oh, Yixuan Li, Kyungwoo Song, Sangdoo Yun, Dongyoon Han

机构 * University of Wisconsin–Madison（威斯康星大学麦迪逊分校） ； Yonsei University（延世大学） ； NAVER AI Lab（NAVER AI实验室） 纠错

Comments ICLR 2025 camera-ready; typo-fixed

BOFormer: Learning to Solve Multi-Objective Bayesian Optimization via Non-Markovian RL

Yu-Heng Hung, Kai-Jie Lin, Yu-Heng Lin, Chien-Yi Wang, Cheng Sun, Ping-Chun Hsieh

机构 * National Yang Ming Chiao Tung University ； NVIDIA Research 纠错

Comments ICLR 2025. Project page and code at https://hungyuheng.github.io/BOFormer/

Federated Granger Causality Learning for Interdependent Clients with State Space Representation

Ayush Mohanty, Nazal Mohamed, Paritosh Ramanan, Nagi Gebraeel

机构 * Georgia Institute of Technology（佐治亚理工学院） ； Oklahoma State University（俄克拉荷马州立大学） 纠错

Comments Published as a conference paper at International Conference on Learning Representations (ICLR) 2025

Nonconvex Stochastic Optimization under Heavy-Tailed Noises: Optimal Convergence without Gradient Clipping

Zijian Liu, Zhengyuan Zhou

机构 * Stern School of Business, New York University（纽约大学斯特恩商学院） 纠错

Comments ICLR 2025

SimBa: Simplicity Bias for Scaling Up Parameters in Deep Reinforcement Learning

Hojoon Lee, Dongyoon Hwang, Donghu Kim, Hyunseung Kim, Jun Jet Tai, Kaushik Subramanian, Peter R. Wurman, Jaegul Choo, Peter Stone, Takuma Seno

机构 * KAIST（韩国科学技术院） ； Sony AI（索尼人工智能） ； KRAFTON ； Coventry University（科文特大学） ； UT Austin（得克萨斯大学奥斯汀分校） 纠错

Comments ICLR'25 (spotlight)

Fast uncovering of protein sequence diversity from structure

Luca Alessandro Silva, Barthelemy Meynard-Piganeau, Carlo Lucibello, Christoph Feinauer

Comments ICLR 2025 spotlight paper

Understanding (Un)Reliability of Steering Vectors in Language Models

Joschka Braun, Carsten Eickhoff, David Krueger, Seyed Ali Bahrainian, Dmitrii Krasheninnikov

机构 * University of Tübingen（图宾根大学） ； Mila, University of Montreal（蒙特利尔大学机器学习研究所） ； University of Cambridge（剑桥大学） 纠错

Comments 17 pages, 10 figures. Presented at the ICLR 2025 Workshop on Foundation Models in the Wild

Enhancing Uncertainty Estimation and Interpretability via Bayesian Non-negative Decision Layer

Xinyue Hu, Zhibin Duan, Bo Chen, Mingyuan Zhou

机构 * National Key Laboratory of Radar Signal Processing, Xidian University（雷达信号处理国家重点实验室，西安电子科技大学） ； School of Mathematics and Statistics, Xi’an Jiaotong University（数学与统计学院，西安交通大学） ； McCombs School of Business, The University of Texas at Austin（商学院，德克萨斯大学奥斯汀分校） 纠错

Comments Accepted by The Thirteenth International Conference on Learning Representations (ICLR 2025)

Taming Transformer Without Using Learning Rate Warmup

Xianbiao Qi, Yelin He, Jiaquan Ye, Chun-Guang Li, Bojia Zi, Xili Dai, Qin Zou, Rong Xiao

机构 * Intellifusion Inc.（Intellifusion公司） ； BUPT（北京邮电大学） ； CUHK（香港中文大学） ； HKUST (GZ)（香港科技大学（广州）） ； WHU（武汉大学） 纠错

Comments This paper is published as a conference paper at ICLR 2025

Simple Guidance Mechanisms for Discrete Diffusion Models

Yair Schiff, Subham Sekhar Sahoo, Hao Phung, Guanghan Wang, Sam Boshar, Hugo Dalla-torre, Bernardo P. de Almeida, Alexander Rush, Thomas Pierrot, Volodymyr Kuleshov

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

Comments ICLR 2025; Code to reproduce our experiments is available here: https://github.com/kuleshov-group/discrete-diffusion-guidance

Diffusion Models as Cartoonists: The Curious Case of High Density Regions

Rafał Karczewski, Markus Heinonen, Vikas Garg

机构 * Aalto University（阿尔托大学） ； YaiYai Ltd（YaiYai公司） 纠错

Comments ICLR 2025

Language Model Alignment in Multilingual Trolley Problems

Zhijing Jin, Max Kleiman-Weiner, Giorgio Piatti, Sydney Levine, Jiarui Liu, Fernando Gonzalez, Francesco Ortu, András Strausz, Mrinmaya Sachan, Rada Mihalcea, Yejin Choi, Bernhard Schölkopf

机构 * Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所） ； ETH Zürich（苏黎世联邦理工学院） ； University of Toronto（多伦多大学） ； University of Washington（华盛顿大学） ； Allen Institute for AI (AI2)（人工智能研究所） ； Carnegie Mellon University（卡内基梅隆大学） ； University of Trieste（特里斯坦大学） ； University of Michigan（密歇根大学） 纠错

Comments ICLR 2025 Spotlight, Best Paper @ NeurIPS 2024 Workshop on Pluralistic Alignment

Decoupled Subgraph Federated Learning

Javad Aliakbari, Johan Östman, Alexandre Graell i Amat

机构 * Chalmers University of Technology（挑战者技术大学） ； AI Sweden（瑞典人工智能） 纠错

Comments This paper has been published at the International Conference on Learning Representations (ICLR), 2025

Journal ref International Conference on Learning Representations (ICLR), 2025

Intrinsic User-Centric Interpretability through Global Mixture of Experts

Vinitra Swamy, Syrielle Montariol, Julian Blackwell, Jibril Frej, Martin Jaggi, Tanja Käser

机构 * EPFL（瑞士联邦理工学院） 纠错

Comments Accepted as a full paper at ICLR 2025 (top 5% of scores) in Singapore

Instructional Segment Embedding: Improving LLM Safety with Instruction Hierarchy

Tong Wu, Shujian Zhang, Kaiqiang Song, Silei Xu, Sanqiang Zhao, Ravi Agrawal, Sathish Reddy Indurthi, Chong Xiang, Prateek Mittal, Wenxuan Zhou

机构 * Princeton University（普林斯顿大学） ； Zoom Video Communications（Zoom视频通讯） 纠错

Comments Preprint

Journal ref ICLR 2025

Can Large Language Models Understand Symbolic Graphics Programs?

Zeju Qiu, Weiyang Liu, Haiwen Feng, Zhen Liu, Tim Z. Xiao, Katherine M. Collins, Joshua B. Tenenbaum, Adrian Weller, Michael J. Black, Bernhard Schölkopf

机构 * Max Planck Institute for Intelligent Systems（马克斯·普朗克智能系统研究所） ； University of Cambridge（剑桥大学） ； MIT（麻省理工学院） 纠错

Comments ICLR 2025 Spotlight (v4: 47 pages, 26 figures, project page: https://sgp-bench.github.io/)

WizardCoder: Empowering Code Large Language Models with Evol-Instruct

Ziyang Luo, Can Xu, Pu Zhao, Qingfeng Sun, Xiubo Geng, Wenxiang Hu, Chongyang Tao, Jing Ma, Qingwei Lin, Daxin Jiang

机构 * Microsoft（微软公司） ； Hong Kong Baptist University（香港 Baptist 大学） 纠错

Comments Large Language model, Code Generation, Code LLMs.This paper has been accepted to ICLR 2024. Please cite the ICLR version

Journal ref The Twelfth International Conference on Learning Representations (ICLR 2024)

WizardLM: Empowering large pre-trained language models to follow complex instructions

Can Xu, Qingfeng Sun, Kai Zheng, Xiubo Geng, Pu Zhao, Jiazhan Feng, Chongyang Tao, Qingwei Lin, Daxin Jiang

Comments large language model, instruction fine-tune

Journal ref The Twelfth International Conference on Learning Representations (ICLR 2024)

Correlating instruction-tuning (in multimodal models) with vision-language processing (in the brain)

Subba Reddy Oota, Akshett Jindal, Ishani Mondal, Khushbu Pahwa, Satya Sai Srinath Namburi, Manish Shrivastava, Maneesh Singh, Bapi S. Raju, Manish Gupta

机构 * Technische Universität Berlin（柏林技术大学） ； IIIT Hyderabad（海得拉巴国家理工学院） ； Univ of Maryland（马里兰大学） ； Rice Univ（Rice 大学） ； Univ of Wisconsin - Madison（威斯康星大学麦迪逊分校） ； Spector Inc（Spector 公司） ； Microsoft（微软公司） 纠错

Comments 30 pages, 22 figures, The Thirteenth International Conference on Learning Representations, ICLR-2025, Singapore. https://openreview.net/pdf?id=xkgfLXZ4e0

Journal ref ICLR-2025, Singapore

Multi-modal brain encoding models for multi-modal stimuli

Subba Reddy Oota, Khushbu Pahwa, Mounika Marreddy, Maneesh Singh, Manish Gupta, Bapi S. Raju

机构 * Technische Universität Berlin（柏林技术大学） ； Rice Univ（莱斯大学） ； Univ of Bonn（波恩大学） ； Spector Inc（Spector公司） ； Microsoft（微软） ； IIIT Hyderabad（海得拉巴国家理工学院） 纠错

Comments 26 pages, 15 figures, The Thirteenth International Conference on Learning Representations, ICLR-2025, Singapore. https://openreview.net/pdf?id=0dELcFHig2

Journal ref ICLR-2025, Sinapore