ICML 2025 accepted papers on Graph learning

175 of the 3,342 papers accepted at ICML 2025 (5.2%) are on Graph learning, compared with 5.4% at ICML 2024. They fall into 4 topics.

See ICML 2025 in the interactive map

Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.

Topics and papers

GNN architectures · 69 papers

Can Classic GNNs Be Strong Baselines for Graph-level Tasks? Simple Architectures Meet Excellence (3 citations)
A General Graph Spectral Wavelet Convolution via Chebyshev Order Decomposition (2 citations)
Adaptive Message Passing: A General Framework to Mitigate Oversmoothing, Oversquashing, and Underreaching (1 citation)
Attention Mechanisms Perspective: Exploring LLM Processing of Graph-Structured Data (1 citation)
Best of Both Worlds: Advantages of Hybrid Graph Sequence Models (1 citation)
GrokFormer: Graph Fourier Kolmogorov-Arnold Transformers (1 citation)
RelGNN: Composite Message Passing for Relational Deep Learning (1 citation)
A Cognac Shot To Forget Bad Memories: Corrective Unlearning for Graph Neural Networks
A Dynamical Systems-Inspired Pruning Strategy for Addressing Oversmoothing in Graph Attention Networks
A Manifold Perspective on the Statistical Generalization of Graph Neural Networks
Aggregation Buffer: Revisiting DropEdge with a New Parameter Block
Balancing Efficiency and Expressiveness: Subgraph GNNs with Walk-Based Centrality

and 57 more in the interactive map.

Graph self-supervised learning · 63 papers

Multi-Domain Graph Foundation Models: Robust Knowledge Transfer via Topology Alignment (2 citations)
Neural Graph Matching Improves Retrieval Augmented Generation in Molecular Machine Learning (2 citations)
Position: Graph Learning Will Lose Relevance Due To Poor Benchmarks (2 citations)
DeFoG: Discrete Flow Matching for Graph Generation (1 citation)
Enhancing the Influence of Labels on Unlabeled Nodes in Graph Convolutional Networks (1 citation)
Graph Generative Pre-trained Transformer (1 citation)
Graph Neural Network Generalization With Gaussian Mixture Model Based Augmentation (1 citation)
GraphGPT: Generative Pre-trained Graph Eulerian Transformer (1 citation)
Implicit degree bias in the link prediction task (1 citation)
Learn Beneficial Noise as Graph Augmentation (1 citation)
Modeling All-Atom Glycan Structures via Hierarchical Message Passing and Multi-Scale Pre-training (1 citation)
A Non-Asymptotic Convergent Analysis for Scored-Based Graph Generative Model via a System of Stochastic Differential Equations

and 51 more in the interactive map.

Knowledge and temporal graphs · 23 papers

and 11 more in the interactive map.

Graph algorithms and clustering · 20 papers

and 8 more in the interactive map.