arxivcs.AIcs.CL2026-07-02
Pre-Flight: A Benchmark for Evaluating Large Language Models on Aviation Operational Knowledge
Large language models (LLMs) are increasingly proposed for aviation business operations, from documentation and training generation to customer facing assistants. General purpose benchmarks do not measure whether a model reasons safely and correctly about aviation specific operat…