Graph Transformers
PulseAugur coverage of Graph Transformers — every cluster mentioning Graph Transformers across labs, papers, and developer communities, ranked by signal.
3 day(s) with sentiment data
-
New method enhances graph representation learning with higher-order topology
Researchers have developed a new method for graph representation learning that incorporates higher-order topology. This approach enriches graph representations with topological information through positional encodings, …
-
Model-to-Data Distillation Enhances Graph Neural Networks
Researchers have introduced a novel method called model-to-data (M2D) distillation for Graph Neural Networks (GNNs). This technique transfers properties learned by complex GNNs, such as fairness and robustness, into the…
-
New GHR Framework Enhances Graph Neural Networks for Long-Range Dependencies
Researchers have introduced Graph Hierarchical Recurrence (GHR), a new framework designed to enhance the capabilities of Graph Neural Networks and Graph Transformers. GHR addresses the fundamental limitation these model…
-
Graph Transformers Pre-training Boosted by Supervised Methods in Biochemistry Research
A new research paper explores pre-training methods for graph transformers specifically within the biochemistry field. The study found that supervised pre-training, utilizing computed properties as labels, yielded the mo…
-
Graph Transformers Enhance Super-Resolution for Detonation Flow Analysis
Researchers have developed a novel graph transformer approach, named SR-GT, for mesh-based super-resolution of reacting flows. This method utilizes a graph-based representation compatible with complex geometries and uns…
-
New hybrid encoding enhances node identification in graph neural networks
A new research paper introduces a hybrid distance-spectral graph positional encoding method designed to improve node identification within graph neural networks and graph Transformers. This approach combines anchor-dist…
-
HyPE-GT framework uses hyperbolic encodings for Graph Transformers
Researchers have introduced HyPE-GT, a novel framework that generates learnable positional encodings in hyperbolic space for Graph Transformers. This approach aims to better capture complex hierarchical relationships wi…
-
Cluster Attention (CLATT) enhances graph machine learning models
Researchers have introduced Cluster Attention (CLATT), a novel approach to enhance graph machine learning. CLATT addresses limitations in Message Passing Neural Networks and Graph Transformers by dividing nodes into clu…
-
Survey paper details fairness challenges in augmented graph learning
A new survey paper, "Fairness in Augmented Graph Learning: A Survey," explores the unique fairness challenges introduced by integrating specialized machine learning techniques into graph learning. The paper, termed Fair…
-
Survey details Graph Transformer architectures, theories, and applications
A new survey paper published on arXiv details the advancements in Graph Transformers (GTs), a type of model that enhances graph neural networks by addressing limitations like over-smoothing and over-squashing. The paper…
-
Graph Transformers for MILPs Limited by 1-WL Test, Study Finds
A new paper characterizes the expressive power of global-attention graph transformers used for mixed-integer linear programs (MILPs). The research proves that these models, including architectures like Graphormer and Se…
-
GenGNN advances discrete graph generation with faster, local message-passing
Researchers have developed GenGNN, a novel message-passing backbone for discrete graph generation that challenges the necessity of Graph Transformers or higher-order architectures. This new model demonstrates strong per…
-
New Graph Convolutional Attention Method Improves Spectral Denoising
Researchers have introduced Graph Convolutional Attention (GCA), a novel method for graph denoising and diffusion that offers a spectral perspective. Unlike standard linear attention, GCA directly utilizes the input gra…
-
Graph Neural Networks applied to optimization and physics problems · 2 sources tracked
Researchers are exploring the application of graph neural networks (GNNs) beyond their traditional roles in combinatorial optimization and theoretical physics. One study demonstrates that GNNs can function as effective …
-
X-LogSMask enhances Transformers for graph data, achieving SOTA on 13 benchmarks
Researchers have developed X-LogSMask, a novel method to adapt Transformer architectures for graph-structured data. This technique injects graph topology directly into attention logits, allowing each attention head to o…
-
New CIPE method enhances Transformer performance on graph data
Researchers have developed a new method called Communicability-Inspired Positional Encoding (CIPE) to improve how Transformers process non-Euclidean graph data. CIPE creates a geometry where inner products reflect struc…
-
Graph Transformers Show Size Transferability, Matching GNNs
Researchers have established a theoretical connection between Graph Transformers (GTs) and Manifold Neural Networks (MNNs), particularly when GTs utilize GNN-based positional encodings. This study demonstrates that GTs …
-
GNNs struggle to approximate sparse matrix factorizations
A new research paper demonstrates that standard message-passing Graph Neural Networks (GNNs) are fundamentally unable to approximate sparse triangular factorizations. The study shows that even advanced architectures lik…
-
New logic-based graph learning method rivals GNNs in speed and performance
Researchers have developed new variants of the Weisfeiler-Leman algorithm for graph classification, which involve modifying the underlying logical framework. These variants allow graph data to be tabularized, enabling t…
-
New GHR framework enhances graph neural networks for long-range dependencies
Researchers have introduced Graph Hierarchical Recurrence (GHR), a new framework designed to improve how Graph Neural Networks and Graph Transformers handle long-range dependencies within graph data. GHR operates on bot…