liveBook · Manning

Graph neural network

As explained in Retrieval Augmented Generation, The Foundational Ideas

A graph neural network is a neural network that learns representations from graph-structured data by aggregating information across connected nodes.

A graph neural network (GNN) refines representations by aggregating information from connected nodes. In retrieval systems, this lets related passages reinforce one another while reducing the influence of noisy or misleading evidence before generation. section 5.6.1

A GNN-based passage-refinement pipeline proceeds in three stages:

  • Construct a graph whose nodes represent passages and whose edges connect closely related passages, such as passages sharing entities. section 5.6.1
  • Aggregate neighboring passage information to strengthen relevant embeddings and dampen noise. section 5.6.1
  • Rerank and prune the refined passages before sending a smaller, denser context to the FiD encoder and decoder. section 5.6.1

KG-FiD

KG-FiD applies a GNN as a pre-filter for multi-document question answering. By removing distracting passages before fusion, it reduces decoder workload and can improve answer quality when retrieval produces redundant or misleading context. section 5.6.1

How Retrieval Augmented Generation, The Foundational Ideas covers it

How deep it goes
Defines neighbor aggregation and gives a three-stage pipeline, but stops short of layer details, training procedures, or message-passing equations.
Examples
Demonstrates KG-FiD with passage nodes, entity-sharing edges, and a curated subset passed to FiD.
What you can do after
Apply it via KG-FiD to pre-filter noisy passages before encoder-decoder fusion.
How it connects
Connects graph construction to entity-linked passages, embedding refinement, passage reranking, pruning, and Fusion-in-Decoder.

Also explained in

Compare how 8 books explain graph neural network →

Related concepts

This explainer was generated by Manning's knowledge graph from the text of Retrieval Augmented Generation, The Foundational Ideas. The section links point to the parts of the book it draws on.