Knowledge Tree
Relations (1)
related 2.58 — strongly supporting 5 facts
KGHaluBench is a benchmark specifically designed to evaluate and identify instances of hallucination in Large Language Models, as described in [1], [2], and [3]. The benchmark utilizes a verification framework to assess factuality and detect various types of hallucinations, as detailed in [4] and [5].
Facts (5)
Sources
A Knowledge Graph-Based Hallucination Benchmark for Evaluating ... arxiv.org 2 facts
procedureThe KGHaluBench response verification framework assesses the factuality of long-form text by identifying hallucinations through three steps: (1) an abstention filter to detect expressions of uncertainty, (2) an initial entity-level filter to identify semantic misalignment with the entity, and (3) a final fact-level check to verify correctness against grounded facts.
referenceTable 1 of the KGHaluBench experiments provides the weighted accuracy, abstain rate, and both hallucination rates for all models tested.
A Knowledge Graph-Based Hallucination Benchmark for Evaluating ... aclanthology.org 2 facts
procedureThe KGHaluBench automated verification pipeline detects abstentions and verifies Large Language Model responses at both conceptual and correctness levels to identify different types of hallucinations.
procedureThe KGHaluBench automated verification pipeline detects abstentions and verifies LLM responses at both conceptual and correctness levels to identify different types of hallucinations.
KGHaluBench: A Knowledge Graph-Based Hallucination ... researchgate.net 1 fact
claimKGHaluBench is a Knowledge Graph-based hallucination benchmark designed to evaluate Large Language Models.