← Return to Article Details A Multimodal and Dynamically Updatable Benchmark for Aviation Question Answering with Large Language Models