Skip to main content

Module explainability

Module explainability 

Source
Expand description

Explainability engine for graph-based RAG — attention weights, path explanation, provenance.

Provides human-interpretable explanations of why a set of triples was retrieved for a given query. Three explanation strategies:

  1. Attention weights — normalized relevance scores per triple
  2. Path explanation — BFS shortest-path from query entity to result entity
  3. Provenance — source documents / IRIs that contributed to the answer

Structs§

AttentionExplanation
Normalized attention weights over a set of triples.
ExplainabilityEngine
Top-level explainability engine — wraps all explanation strategies.
PathExplanation
A BFS shortest-path explanation from a query entity to a result entity.
PathHop
One hop in a graph path.
ProvenanceReport
Source provenance for a set of triples.
ScoredTriple
A knowledge graph triple with an associated relevance score.