RAG over textual graphs: retrieving a query-relevant subgraph with a soft pruner, feeding the subgraph's text and structure to the LLM, and results on WebQSP and ExplaGraphs.
← all seminars & paper readings