Graph Transformers
PulseAugur coverage of Graph Transformers — every cluster mentioning Graph Transformers across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
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…
-
Graph Transformers training issues identified and adaptive control proposed
Researchers have identified a phenomenon called distance-misaligned training in Graph Transformers, where the model's communication allocation doesn't match the location of relevant information for a given task. They de…