liveBook · Manning

Permutation invariance

As explained in Build AI Drug Discovery Pipelines

Permutation invariance is the property that a computation produces the same result when its unordered inputs are presented in a different order.

Permutation invariance means producing the same result when unordered inputs appear in a different order. In a graph neural network, a node’s neighbors have no natural ordering, so their messages must be combined in an order-independent way. Summing or averaging preserves the result even when graph numbering changes. section 11.4

Within message passing, permutation invariance supports:

  • Constructing each message from a neighboring node. section 11.4
  • Aggregating messages with a sum or mean. section 11.4
  • Updating the node representation from the aggregate and its current value. section 11.4
hi(k+1) = σ( W(k) AGG( { hj(k) : j ∈ N(i) ∪ {i} } ) )

A graph convolution illustrates permutation-invariant aggregation by combining a node with its neighbors before applying a learned transformation and nonlinearity. section 11.4

How Build AI Drug Discovery Pipelines covers it

How deep it goes
Defines the requirement, explains sum and mean aggregation, and connects it to message passing while deferring full derivations.
Examples
Uses atom and residue neighborhoods, sum or mean aggregation, and PyTorch Geometric edge_index, x, and edge_attr tensors.
What it assumes
Assumes node features, graph edges, and neighbor relationships while explaining their roles in message passing.
What you can do after
Evaluate/understand why graph models return consistent node representations when unordered neighbors are renumbered.
How it connects
Connects message passing, graph neural networks, node features, edge features, graph convolution, and PyTorch Geometric.

Also explained in

Compare how 4 books explain permutation invariance →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Build AI Drug Discovery Pipelines. The section links point to the parts of the book it draws on.