Granola: Adaptive normalization for graph neural networks

M Eliasof, B Bevilacqua, CB Schönlieb… - arxiv preprint arxiv …, 2024‏ - arxiv.org
In recent years, significant efforts have been made to refine the design of Graph Neural
Network (GNN) layers, aiming to overcome diverse challenges, such as limited expressive …

[PDF][PDF] Digraf: Diffeomorphic graph-adaptive activation function

KSI Mantri, X Wang, CB Schönlieb… - arxiv preprint arxiv …, 2024‏ - proceedings.neurips.cc
In this paper, we propose a novel activation function tailored specifically for graph data in
Graph Neural Networks (GNNs). Motivated by the need for graph-adaptive and flexible …

Enhancing High-order Interaction Awareness in LLM-based Recommender Model

X Wang, J Cui, F Fukumoto, Y Suzuki - arxiv preprint arxiv:2409.19979, 2024‏ - arxiv.org
Large language models (LLMs) have demonstrated prominent reasoning capabilities in
recommendation tasks by transforming them into text-generation tasks. However, existing …

Graphmaker: Can diffusion models generate large attributed graphs?

M Li, E Kreačić, VK Potluru, P Li - arxiv preprint arxiv:2310.13833, 2023‏ - arxiv.org
Large-scale graphs with node attributes are increasingly common in various real-world
applications. Creating synthetic, attribute-rich graphs that mirror real-world examples is …

NT-LLM: A Novel Node Tokenizer for Integrating Graph Structure into Large Language Models

Y Ji, C Liu, X Chen, Y Ding, D Luo, M Li, W Lin… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Graphs are a fundamental data structure for representing relationships in real-world
scenarios. With the success of Large Language Models (LLMs) across various natural …

Motif-driven molecular graph representation learning

R Wang, Y Ma, X Liu, Z **ng, Y Shen - Expert Systems with Applications, 2025‏ - Elsevier
Abstract Graph Neural Networks (GNNs) have emerged as powerful tools for molecular
graph analysis. Subgraph-based GNNs focus on learning high-level local patterns beyond …

Learning Efficient Positional Encodings with Graph Neural Networks

CI Kanatsoulis, E Choi, S Jegelka, J Leskovec… - arxiv preprint arxiv …, 2025‏ - arxiv.org
Positional encodings (PEs) are essential for effective graph representation learning because
they provide position awareness in inherently position-agnostic transformer architectures …

Towards Graph Foundation Models: A Study on the Generalization of Positional and Structural Encodings

BJ Franks, M Eliasof, S Cantürk, G Wolf… - arxiv preprint arxiv …, 2024‏ - arxiv.org
Recent advances in integrating positional and structural encodings (PSEs) into graph neural
networks (GNNs) have significantly enhanced their performance across various graph …

On the Effectiveness of Random Weights in Graph Neural Networks

T Bui, CB Schönlieb, B Ribeiro, B Bevilacqua… - arxiv preprint arxiv …, 2025‏ - arxiv.org
Graph Neural Networks (GNNs) have achieved remarkable success across diverse tasks on
graph-structured data, primarily through the use of learned weights in message passing …

LASE: Learned Adjacency Spectral Embeddings

SP Casulo, M Fiori, F Larroca, G Mateos - arxiv preprint arxiv:2412.17734, 2024‏ - arxiv.org
We put forth a principled design of a neural architecture to learn nodal Adjacency Spectral
Embeddings (ASE) from graph inputs. By bringing to bear the gradient descent (GD) method …