Concept-Based Unsupervised Domain Adaptation
Comments Accepted by ICML 2025
期刊&会议
ICML
International Conference on Machine Learning · 会议 · Machine Learning
When Bad Data Leads to Good Models
Comments ICML 2025
Communicating Activations Between Language Model Agents
Comments ICML 2025
Look Twice Before You Answer: Memory-Space Visual Retracing for Hallucination Mitigation in Multimodal Large Language Models
Comments Accepted by ICML 2025
Retraining with Predicted Hard Labels Provably Increases Model Accuracy
Comments To appear in ICML 2025
A Probabilistic Approach to Learning the Degree of Equivariance in Steerable CNNs
Comments 9 pages, published at ICML 2024 as main conference paper
Position: AI Evaluation Should Learn from How We Test Humans
Comments Accepted by ICML 2025
On Path to Multimodal Generalist: General-Level and General-Bench
Comments ICML'25, 305 pages, 115 tables, 177 figures, project page: https://generalist.top/
Vision-Language Models Create Cross-Modal Task Representations
机构 * University of California, Berkeley, USA（加州大学伯克利分校）
Comments ICML 2025
Learning Survival Distributions with the Asymmetric Laplace Distribution
机构 * Department of Electrical and Computer Engineering, Duke University（电子工程系，杜克大学）
Comments Accepted to ICML 2025
Grokking in the Wild: Data Augmentation for Real-World Multi-Hop Reasoning with Transformers
Comments Accepted to the International Conference on Machine Learning (ICML) 2025
Maximum Coverage in Turnstile Streams with Applications to Fingerprinting Measures
Comments ICML 2025. Added experiments in V2
Liger: Linearizing Large Language Models to Gated Recurrent Structures
Comments Accepted by ICML 2025, 15 pages
Vision-Language Model Selection and Reuse for Downstream Adaptation
Comments Accepted by ICML 2025
SUICA: Learning Super-high Dimensional Sparse Implicit Neural Representations for Spatial Transcriptomics
机构 * Qingtian Zhu（朱庆天） ； Yumin Zheng（郑玉敏） ； Yuling Sang（Sang Yuling） ； Yifan Zhan（湛毅凡） ； Ziyan Zhu（朱子燕） ； Jun Ding（丁俊） ； Yinqiang Zheng（郑延庆）
Comments ICML 2025
Enhancing Statistical Validity and Power in Hybrid Controlled Trials: A Randomization Inference Approach with Conformal Selective Borrowing
Comments Accepted by ICML 2025
Boosting Masked ECG-Text Auto-Encoders as Discriminative Learners
Comments Accepted at ICML 2025
On Understanding Attention-Based In-Context Learning for Categorical Data
Comments Accepted to ICML 2025
Transport meets Variational Inference: Controlled Monte Carlo Diffusions
机构 * University of Cambridge（剑桥大学） ； KIT Karlsruhe（卡尔斯鲁厄理工学院） ； Kings College London（伦敦国王学院）
Comments Workshop on New Frontiers in Learning, Control, and Dynamical Systems at the International Conference on Machine Learning (ICML), Honolulu, Hawaii, USA, 2023
Tight Regret Bounds for Bayesian Optimization in One Dimension
机构 * Jonathan Scarlett（独立研究者）
Comments ICML 2018 + supplementary material. This version also includes an 'Errata' section correcting two minor mistakes
Actor-Critics Can Achieve Optimal Sample Efficiency
Comments Accepted to ICML 2025
Synthesizing Images on Perceptual Boundaries of ANNs for Uncovering and Manipulating Human Perceptual Variability
Comments accepted at ICML 2025
Ergodic Generative Flows
Comments 20 pages, 5 figures, 1 table, accepted at ICML 2025
Advancing Constrained Monotonic Neural Networks: Achieving Universal Approximation Beyond Bounded Activations
Comments International Conference on Machine Learning
UI-Vision: A Desktop-centric GUI Benchmark for Visual Perception and Interaction
机构 * Mila - Quebec AI Institute（魁北克AI研究所） ； University of Waterloo（滑铁卢大学） ； National University of Singapore（新加坡国立大学） ； CIFAR AI Chair（CIFAR人工智能职位）
Comments This paper has been accepted to the 41st International Conference on Machine Learning (ICML 2025)
Improving Model Alignment Through Collective Intelligence of Open-Source LLMS
机构 * Duke University（杜克大学） ； Together AI ； University of Chicago（芝加哥大学） ； Stanford University（斯坦福大学）
Comments ICML 2025
Radio: Rate-Distortion Optimization for Large Language Model Compression
Comments Accepted to ICML 2025
A Sharper Global Convergence Analysis for Average Reward Reinforcement Learning via an Actor-Critic Approach
Comments Accepted to the 42nd International Conference on Machine Learning (ICML), 2025
SIMPLEMIX: Frustratingly Simple Mixing of Off- and On-policy Data in Language Model Preference Learning
机构 * Johns Hopkins University（约翰霍普金斯大学）
Comments To appear in ICML 2025
Exogenous Isomorphism for Counterfactual Identifiability
Comments 43 pages, 4 figures. Accepted at ICML 2025 (Spotlight poster)
