liveBook · Manning

KG-FiD

As explained in Retrieval Augmented Generation, The Foundational Ideas

KG-FiD is a Fusion-in-Decoder extension that uses graph-based passage refinement before decoding.

KG-FiD addresses a common Fusion-in-Decoder failure: noisy, redundant, or misleading passages distract the decoder and leave relevant evidence unextracted. It refines passages before standard FiD encoding and decoding. section 5.6.1

Its pipeline uses three stages:

  • Graph construction connects passage nodes with edges when passages are closely related, such as when they share entities. section 5.6.1
  • GNN refinement updates each passage embedding by aggregating information from neighboring passages, strengthening relevant evidence and dampening noise. section 5.6.1
  • Reranking and pruning use the refined embeddings to select a smaller, curated passage set for the standard FiD encoder and decoder. section 5.6.1

Entity-linked passage refinement

When retrieved passages share entities, graph edges let neighboring evidence influence passage embeddings before reranking. The decoder then receives a denser context with fewer distracting passages. section 5.6.1

How Retrieval Augmented Generation, The Foundational Ideas covers it

How deep it goes
Explains graph construction, GNN refinement, reranking, and pruning, but stops short of implementation details.
Examples
Uses passage graphs whose edges represent closely related passages, including passages sharing entities.
What it assumes
Assumes familiarity with passage retrieval and FiD while defining graph nodes, edges, GNN refinement, reranking, and pruning.
What you can do after
Build it from scratch as a three-stage passage-refinement pipeline feeding a standard FiD encoder and decoder.
How it connects
Connects Fusion-in-Decoder with graph neural networks, passage reranking, and context pruning.

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.