VArify: A Visual Analytics System for Verifying Knowledge Enhanced Large Language Model Responses in Food Science
Graph Retrieval-Augmented Generation (GraphRAG) enables Large Language Models (LLMs) to leverage structured, domain-specific knowledge graph databases for factually grounded responses. However, the retrieval of irrelevant or conflicting data can still result in erroneous responses. In knowledge-intensive and evidence-focused domains, human verification of the supporting evidence for an LLM response is still necessary. We conducted a formative pilot study to characterize the challenges of verifying complex, multi-layered data retrieved by GraphRAG systems. Based on these insights, we present VArify, a visual analytics system that leverages a file directory-inspired tree visualization to support simultaneous exploration of inter-group relationships and intra-group hierarchies within the retrieved evidence. We evaluate VArify through a user study with six food science experts and students. Our results indicate that the system effectively helps users distinguish between an LLM's internal parametric knowledge and external graph-sourced evidence. Furthermore, the visualization helped experts identify inaccuracies within the underlying knowledge graph itself, leading to more calibrated trust in the model's output. We conclude by discussing opportunities to leverage visualizations to further support verification regarding unknown unknowns, personalization, and limitations of knowledge graphs.