ICML 2025

3,384 papers

Top Research Directions

Research directions most represented by ICML 2025 papers.

Shifting from Static Benchmarks to Dynamic Evaluation 37
Inference Time Compute Scaling for Complex Reasoning 30
Clean Accuracy Degradation in Machine Learning Defenses 29
Overcoming the Iterative Inference Latency of Diffusion Models 27
The Performance Cliff in Ultra Low Bit Quantization 27
The Computational Overhead of Generative Data Synthesis 23
Bypassing the Computational Bottleneck of Adversarial Training 23
Outlier Suppression for Low Bit Transformer Quantization 20
The Function Approximation Gap in Reinforcement Learning Theory 20
Understanding Training Data Memorization in Generative Models 18
Resolving Superposition with Sparse Autoencoders 17
Preserving Model Utility During Machine Unlearning 17
Adapting Graph Neural Networks to Evolving Topologies 17
Overcoming the Homophily Assumption in Graph Neural Networks 16
Over-Smoothing in Graph Neural Networks 16

Top Authors

AuthorPapers
Mihaela van der Schaar15
Dacheng Tao12
Guancheng Wan12
Li Shen12
Mang Ye12
Wenke Huang12
Sergey Levine11
Kun Zhang10
Quanquan Gu10
Sanmi Koyejo10
Bo Han9
Feng Liu9
Ion Stoica9
Junchi Yan9
Martin Vechev9
Ruixuan Li9
Shirui Pan9
Yu Cheng9
Jiang Bian8
Jun Zhu8

Impactful Papers

Papers from ICML 2025 that shaped research clusters.

Foundational

Influential

Papers

Title
Understanding Complexity in VideoQA via Visual Program Generation
Understanding Fixed Predictions via Confined Regions
Understanding Generalization in Quantum Machine Learning with Margins
Understanding High-Dimensional Bayesian Optimization
Understanding Input Selectivity in Mamba: Impact on Approximation Power, Memorization, and Associative Recall Capacity
Understanding Mode Connectivity via Parameter Space Symmetry
Understanding Model Ensemble in Transferable Adversarial Attack
Understanding Model Reprogramming for CLIP via Decoupling Visual Prompts
Understanding Multimodal LLMs Under Distribution Shifts: An Information-Theoretic Approach
Understanding Nonlinear Implicit Bias via Region Counts in Input Space
Understanding Overadaptation in Supervised Fine-Tuning: The Role of Ensemble Methods
Understanding Sharpness Dynamics in NN Training with a Minimalist Example: The Effects of Dataset Difficulty, Depth, Stochasticity, and More
Understanding Synthetic Context Extension via Retrieval Heads
Understanding the difficulties of posterior predictive estimation
Understanding the Emergence of Multimodal Representation Alignment
Understanding the Forgetting of (Replay-based) Continual Learning via Feature Learning: Angle Matters
Understanding the Kronecker Matrix-Vector Complexity of Linear Algebra
Understanding the Limits of Deep Tabular Methods with Temporal Shift
Understanding the Logic of Direct Preference Alignment through Logic
Understanding the Skill Gap in Recurrent Language Models: The Role of the Gather-and-Aggregate Mechanism
Understanding the Statistical Accuracy-Communication Trade-off in Personalized Federated Learning with Minimax Guarantees
Understanding the Unfairness in Network Quantization
UnHiPPO: Uncertainty-aware Initialization for State Space Models
UniDB: A Unified Diffusion Bridge Framework via Stochastic Optimal Control
Unifews: You Need Fewer Operations for Efficient Graph Neural Networks
Unified Analysis of Continuous Weak Features Learning with Applications to Learning from Missing Data
Unified Breakdown Analysis for Byzantine Robust Gossip
Unified K-Means Clustering with Label-Guided Manifold Learning
Unified Screening for Multiple Diseases
Uniform Mean Estimation for Heavy-Tailed Distributions via Median-of-Means
Unifying 2D and 3D Vision-Language Understanding
Unifying Knowledge from Diverse Datasets to Enhance Spatial-Temporal Modeling: A Granularity-Adaptive Geographical Embedding Approach
Unifying Specialized Visual Encoders for Video Language Models
UniMate: A Unified Model for Mechanical Metamaterial Generation, Property Prediction, and Condition Confirmation
UniMC: Taming Diffusion Transformer for Unified Keypoint-Guided Multi-Class Image Generation
UniMoMo: Unified Generative Modeling of 3D Molecules for De Novo Binder Design
UniSim: A Unified Simulator for Time-Coarsened Dynamics of Biomolecules
Unisolver: PDE-Conditional Transformers Towards Universal Neural PDE Solvers
Unisoma: A Unified Transformer-based Solver for Multi-Solid Systems
Universal Approximation of Mean-Field Models via Transformers
Universal Approximation Theorem of Deep Q-Networks
Universal Biological Sequence Reranking for Improved De Novo Peptide Sequencing
Universal Length Generalization with Turing Programs
Universal Neural Optimal Transport
Universal Sparse Autoencoders: Interpretable Cross-Model Concept Alignment
Unlocking Post-hoc Dataset Inference with Synthetic Data
Unlocking the Capabilities of Large Vision-Language Models for Generalizable and Explainable Deepfake Detection
Unlocking the Power of Rehearsal in Continual Learning: A Theoretical Perspective
Unlocking the Power of SAM 2 for Few-Shot Segmentation
unMORE: Unsupervised Multi-Object Segmentation via Center-Boundary Reasoning

« prev Page 65/68 next »