reference
Research directions for hallucination evaluation include the development of integrated, multi-task, multilingual benchmarks with unified annotation schemas (Liu et al., 2023) and the use of attribution-aware metrics incorporating Prompt Sensitivity (PS) and Model Variability (MV).
Authors
Sources
- Survey and analysis of hallucinations in large language models www.frontiersin.org via serper
Referenced by nodes (3)
- Prompt Sensitivity concept
- Model Variability concept
- Hallucination Evaluation Model concept