measurement
The correlation between actual and estimated question difficulty in KGHaluBench is moderate and negative, with a Spearman's rank correlation coefficient of -0.403 and a Kendall's rank correlation coefficient of -0.299.
Authors
Sources
- A Knowledge Graph-Based Hallucination Benchmark for Evaluating ... arxiv.org via serper
Referenced by nodes (1)
- KGHaluBench concept