InfoResearchPreprintLLM-specific
Measuring Hallucination in Large Language Models for Cyber Threat Intelligence: An Exploratory Study
- Published
- Record updated
Summary
Researchers measured hallucinations in LLM-generated cyber threat intelligence outputs using HalluVision, a reproducible framework. They evaluated outputs from two commercial LLMs and one open-weight LLM across 4,940 real-world cybersecurity articles, validating results against 100 human-annotated outputs and two faithfulness detectors. Semantic similarity aligned with human judgments, while entity-based metrics showed limited agreement.