Applies the Softmax function to an n-dimensional input Tensor rescaling them so that the elements of the n-dimensional output Tensor lie in the range [0,1] and sum to 1. g., Dynamic Edge-Conditioned Filters in Convolutional Networks on Graphs paper, which overlays a regular grid of user-defined size over a point cloud and clusters all points within the same voxel. Memory based pooling layer from "Memory-Based Graph Networks" paper, which learns a coarsened graph representation based on soft cluster assignments. The RotatE model from the "RotatE: Knowledge Graph Embedding by Relational Rotation in Complex Space" paper. Creates a criterion that optimizes a multi-label one-versus-all loss based on max-entropy, between input x x x and target y y y of size ( N , C ) (N, C) ( N , C ).

The crystal graph convolutional operator from the "Crystal Graph Convolutional Neural Networks for an Accurate and Interpretable Prediction of Material Properties" paper. The MetaPath2Vec model from the "metapath2vec: Scalable Representation Learning for Heterogeneous Networks" paper where random walks based on a given metapath are sampled in a heterogeneous graph, and node embeddings are learned via negative sampling optimization. The powermean aggregation operator based on a power term, as described in the "DeeperGCN: All You Need to Train Deeper GCNs" paper. The chebyshev spectral graph convolutional operator from the "Convolutional Neural Networks on Graphs with Fast Localized Spectral Filtering" paper.The soft attention aggregation layer from the "Graph Matching Networks for Learning the Similarity of Graph Structured Objects" paper.

The dynamic neighborhood aggregation operator from the "Just Jump: Towards Dynamic Neighborhood Aggregation in Graph Neural Networks" paper.Rearranges elements in a tensor of shape ( ∗ , C × r 2 , H , W ) (*, C \times r Applies instance normalization over each individual example in a batch of node features as described in the "Instance Normalization: The Missing Ingredient for Fast Stylization" paper. BatchNorm2d module with lazy initialization of the num_features argument of the BatchNorm2d that is inferred from the input. g., the j j j-th channel of the i i i-th sample in the batched input is a 1D tensor input [ i , j ] \text{input}[i, j] input [ i , j ]).

Notably, all aggregations share the same set of forward arguments, as described in detail in the torch_geometric.InstanceNorm2d module with lazy initialization of the num_features argument of the InstanceNorm2d that is inferred from the input. The Gini coefficient from the "Improving Molecular Graph Neural Network Explainability with Orthonormalization and Induced Sparsity" paper. The DimeNet++ from the "Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Molecules" paper. First, aggregations can be resolved from pure strings via a lookup table, following the design principles of the class-resolver library, e.

Allows the model to jointly attend to information from different representation subspaces as described in the paper: Attention Is All You Need. The PointNet set layer from the "PointNet: Deep Learning on Point Sets for 3D Classification and Segmentation" and "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space" papers. The Graph Neural Network from the "Dynamic Graph CNN for Learning on Point Clouds" paper, using the EdgeConv operator for message passing.The graph convolutional operator with initial residual connections and identity mapping (GCNII) from the "Simple and Deep Graph Convolutional Networks" paper. The ClusterGCN graph convolutional operator from the "Cluster-GCN: An Efficient Algorithm for Training Deep and Large Graph Convolutional Networks" paper. The k-NN interpolation from the "PointNet++: Deep Hierarchical Feature Learning on Point Sets in a Metric Space" paper. The continuous-filter convolutional neural network SchNet from the "SchNet: A Continuous-filter Convolutional Neural Network for Modeling Quantum Interactions" paper that uses the interactions blocks of the form. Performs MLP aggregation in which the elements to aggregate are flattened into a single vectorial representation, and are then processed by a Multi-Layer Perceptron (MLP), as described in the "Graph Neural Networks with Adaptive Readouts" paper.